MicroRNA Interactome Multiomics Characterization for Cancer Research and Personalized Medicine: An Expert Review

Abstract

MicroRNAs (miRNAs) that are mutually modulated by their interacting partners (interactome) are being increasingly noted for their significant role in pathogenesis and treatment of various human cancers. Recently, miRNA interactome dissected with multiomics approaches has been the subject of focus since individual tools or methods failed to provide the necessary comprehensive clues on the complete interactome. Even though single-omics technologies such as proteomics can uncover part of the interactome, the biological and clinical understanding still remain incomplete. In this study, we present an expert review of studies involving multiomics approaches to identification of miRNA interactome and its application in mechanistic characterization, classification, and therapeutic target identification in a variety of cancers, and with a focus on proteomics. We also discuss individual or multiple miRNA-based interactome identification in various pathological conditions of relevance to clinical medicine. Various new single-omics methods that can be integrated into multiomics cancer research and the computational approaches to analyze and predict miRNA interactome are also highlighted in this review. In all, we contextulize the power of multiomics approaches and the importance of the miRNA interactome to achieve the vision and practice of predictive, preventive, and personalized medicine in cancer research and clinical oncology.

Introduction

There has been a significant increase in the number of studies on the role of miRNAs in normal physiological processes and diseases, including cancer (Peng and Croce, 2016). Both individual and group of miRNAs are known to regulate the progression of carcinogenesis, metastasis, and tumor microenvironment (Zaheer et al., 2019). MiRNAs targeting oncogenes are termed as tumor suppressor miRNAs, and those targeting tumor suppressors are termed as oncomiRNAs (Svoronos et al., 2016). They are being increasingly used as biomarkers (Isenovic et al., 2017), and in miRNA replacement or inhibition therapies (Hosseinahli et al., 2018; Mollaei et al., 2019; Nguyen and Chang, 2018). MiRNAs play an important role in the therapeutic resistance of cancers and many are currently in various phases of clinical trials (Guo et al., 2020; Johora et al., 2019; Ors-kumoglu et al., 2019; Si et al., 2019).

The growing importance of miRNAs in cancer research and personalized medicine (Akyay et al., 2020) led to investigation of miRNA interactome using single omics followed by multiomics. Multiomics enables identification/characterization of biomarkers, disease progression, and treatment response, and more recently is being used to characterize miRNA interactome. This review elaborates on various studies involving multiomics approaches in identifying miRNA interactome using cancer models, with special mention on proteomics technologies and the new single-omics technologies. It also discusses the upcoming challenges, possible solutions, and prospects. Since exosomal and circulatory miRNAs and their roles in cancer are covered in several reviews independently (Bhome et al., 2018; Cui et al., 2019; Kamal and Shahidan, 2020; Sun et al., 2018; Wang et al., 2018), they are excluded in this review.

MiRNA Interactome

MicroRNAs (miRNAs) are small noncoding RNAs (18–22 nucleotides) discovered in Caenorhabditis elegans (Lee et al., 1993) and their biogenesis involves initial transcription as pri-miRNAs, processing by Drosha/DGCR8 (DiGeorge syndrome critical region 8) enzyme complex to pre-miRNAs in the nucleus, later exported by exportin-5, and finally processed into mature miRNAs by the enzyme dicer 1 in the cytoplasm (Denli et al., 2004; Han et al., 2004; Lee et al., 2003; Zhang et al., 2004). Mature miRNA strands are then loaded into RNA-induced silencing complex (RISC) to target the specific protein-coding messenger RNAs (mRNAs). Based on the complementarities between miRNAs and mRNAs, the mRNAs are either degraded (completely or partially) or translationally inhibited (Brien et al., 2018; Kawamata and Tomari, 2010).

MiRNAs not only interact with their target mRNAs but also partner with long noncoding RNAs (lncRNAs, >200 nucleotides in length) and pseudogenes (nonfunctional genes), through competitive binding, resulting in the disruption of several signaling pathways such as angiogenesis and stem cell development (López-urrutia et al., 2019; Sen et al., 2014). In fact, they form a vast regulatory network interacting with the promoter/enhancer regions and transcription factors of target/host genes and modulate gene expression negatively in most cases or even positively in some cases (Tan et al., 2019). MiRNAs also interact with RNA binding proteins (RBPs), including Drosha and Dicer 1 and other RBPs specific to a given miRNA, and inhibit their binding to argonaute 2 (AGO2) protein.

Alternatively, miRNAs could modulate the RBP genes and contribute to the diversity in physiological and pathological processes (Chen et al., 2019; Ciafrè and Galardi, 2013; Zealy et al., 2017). Recent studies have highlighted the importance of polymorphisms in miRNAs and their interacting partners in physiology and pathology (Hrovatin and Kunej, 2018; Piletic and Kunej, 2018).

Figure 1 depicts the miRNA interactome and these interactions can be co-operative or competitive in nature, for example, PUM1 promotes stronger binding of miR-221 and miR-222 to CDKN1B 3′ UTR, promoting its degradation (Kedde et al., 2010).

FIG. 1.

MiRNA interactome: the figure depicts miRNA interacting partners and style of interaction whether miRNAs can act co-operatively or competitively with their partners.

Bioinformatics and Experimental Tools to Identify miRNA Interactome

There are many bioinformatics tools to predict the targets for miRNAs (Ab Mutalib et al., 2019). However, they predict targets based on seed sequence complementarity, free energy, and site conservation status, which are insufficient to predict the total interacting partners of miRNAs as they often end up with significant false positives and/or false negatives; therefore, experimental validation is necessary and this can be accumulated only through single-omics experiments.

MiRNA-mRNA pairs have been identified using reporter assays, cross-linking immunoprecipitation (CLIP), affinity pull-down, degradome sequencing, or bioinformatics prediction (German et al., 2008; Hafner et al., 2010; Helwak et al., 2013; Keene et al., 2006; Kern et al., 2019; Licatalosi et al., 2008; Lu and Leslie, 2016; Moore et al., 2015; Willmann et al., 2014; Yoon et al., 2012). For a specific miRNA, it can be achieved by gene expression profiling after overexpression or inhibition, immunoprecipitation (IP) of RISC with a specific antibody, and pull-down with labeled miRNA mimics (Li and Zhang, 2019).

MiRNA interaction with lncRNAs is identified using CLIP (Kazimierczyk et al., 2020) and enhanced cross-linking and immunoprecipitation (eCLIP), which can additionally identify RBPs regulating miRNAs (Nussbacher and Yeo, 2018). Ribosome profiling is another method that identifies miRNAs associated with ribosomes and unravels changes in protein levels of miRNA targets mediated by translational repression or mRNA degradation and their regulators (Bazzini et al., 2012; Cottrell et al., 2017; Eichhorn et al., 2010).

Expression profiles of miRNAs can be quantified using microarray (Li and Zhang, 2019). Pull-down assays utilize biotinylated miRNAs to identify their partners being pulled down using streptavidin beads and later sequenced. IP uses antibodies against proteins in RISC, particularly argonaute proteins (AGO), to identify mRNAs bound to a miRNA. Modifications include photo-crosslinking of bound RNAs with AGO, such as in CLIP or photoactivatable ribonucleoside-enhanced cross-linking and immunoprecipitation (PAR-CLIP) and cross-linking, linking, and sequencing of hybrids (CLASH is a modification of CLIP method, including sequencing of hybrids) (Li and Zhang, 2019). Degradome sequencing identifies miRNA cleavage sites leading to novel target identification. Figure 2 shows the various tools to identify miRNA interactome and the type of interaction detected by these tools.

FIG. 2.

Tools to characterize miRNA interactome: the figure enlists the various methodologies and techniques used to identify each aspect of the interactome and the type of interactions detected by these tools. The inset A shows direct interaction of miRNA to target leading to its degradation (in black square) as indicated by the arrow, while inset B shows how miRNA directly interacts with RBP 1 (black circle), but the free mRNA binds to other RBP 2 (in dotted square) in the cell, making this RBP 2 an indirect interacting partner of miRNA and alters the cellular response.

Microarray reflects only on the changes in the target mRNAs from a single or multiple miRNA(s), but miss out on changes mediated by translational repression. Similarly, other tools like IP, CLIP, and its associated modifications identify direct interactions between miRNAs, isomiRNAs (isomiRs are variations of miRNAs that are generated from a single miRNA precursor locus), and their partners, that is, they identify targets regulated by both translational repression and mRNA degradation, but leave out the indirect interactions. The results from these experimental studies provide the basis for the single-omics data, that is, global measurement of the subset of cellular processes or molecules.

Single Omics as a Tool to Study miRNA Interactome

Single omics have long been used to identify biomarkers, differentially expressed genes, proteins, miRNAs, and miRNA targets to distinguish between different physiological and pathological states of the cell. Genomics delves into the organism's genome, particularly the structure, function, and mapping of genes, whereas epigenomics deals with changes in the DNA without changes in its sequence such as associated histone modifications and DNA methylations. Transcriptomics exhibits the transcriptional status of the cell at specific circumstances, including protein-coding mRNAs and noncoding RNAs and their associated properties. Proteomics reveals information about the various cellular protein levels and their post-translational modifications, and metabolomics divulges details about the small molecules called metabolites derived from cellular processes.

Individual omics studies only one part of the miRNA interactome, such as genomics would reveal changes in the miRNA gene that affect its expression; epigenomics would reveal epigenetic factors influencing miRNA expression or vice versa; transcriptomics would reveal the overall RNA targetome of the miRNA interactome; and unlike the rest, proteomics would quantify miRNA-induced changes at the protein level studying both direct and indirect interactions.

Many techniques have been developed to label and quantify proteins (Giambruno et al., 2018). Various techniques involving mass spectrometry (MS)-based proteomics, with modifications such as affinity purification, pull-down assays, and hybridization-based purification, are known (Giambruno et al., 2018). Untargeted proteomics involves covering as much of the proteome as possible, while targeted proteomics aims to target a specific subset of the proteome to improve specificity and sensitivity, which can be used for developing biomarkers and to unravel the details on the targetome of miRNA(s) (Nouri-Nigjeh et al., 2015).

To further improve single-omics methods to detect and quantify miRNAs, several developments have been made and shown in Table 1. The advancements in single omics are mainly an improvement over quantitative real-time PCR (qRT-PCR) to detect and quantify miRNAs. Validation of omics from microarray is done using qRT-PCR and although it is quite sensitive, there are still some disadvantages such as lack of specificity and inability to distinguish between genome location of miRNA host genes and their paralogs, and is labor- and cost-intensive (Forero et al., 2019). Lack of specificity and labor-intensiveness has been overcome by MS-based proteomics (Xu et al., 2015) and it converts miRNA signal into reporter peptide signal, which is later quantified by MS using a peptide substrate that is complementary to target miRNA.

Table 1.

Single Omics Advances in the Context of Cancers

Objective(S)	Techniques	Model	Software	Outcome	Reference
To quantify miR-21	LC-MS/MS-based quasi-targeted proteomics assay	MCF-10A, the parental drug-sensitive normal breast epithelial cells MCF-7/WT and the drug-resistant cancer cells MCF-7/ADR, and 36 pairs of human breast primary tumors and adjacent normal tissue samples.	STEPP software, ESP predictor	A novel LC-MS/MS-based quasi-targeted proteomics assay was developed and validated for quantification of miR-21 in biological samples, including 3 breast cell lines.	Xu et al., (2015)
To quantify miRNAs and isomiRs and their modifications	Small RNA sequencing method modified to quantify miRNAs and isomiRs	HEK293T and HeLa	N/A	Identification of isomiRs for use in diagnosis.	Kim et al., (2019)
To multiplex and quantify miRNA from the most commonly available biological samples without additional sample processing	Method to spatially resolve and quantify miRNAs in FFPE tissues	KrasLSL-G12D/+; p53fl/fl genetically engineered mouse model for non-small cell lung cancer	N/A	Quantifies up to 9 miRNAs directly from FFPE tissues without amplification, thereby avoiding sequence bias. Minimal processing steps allowing quantification in a rapid manner. Disadvantage: inability to quantify miRNAs that are expressed at extremely low levels.	Nagarajan et al., (2020)
To identify miRNA targets	O16/O18 labeling for LC-MS-based quantitative proteomics	Kaposi sarcoma-associated herpes virus (KSHV) miR-K1 target prediction; HEK 293T miR-K1 vs. empty vector.	Kullback–Leiberler distance test (KL test), TPP for peptide identification	Developed and applied a proteomics approach for identifying the targets of KSHV miRNA. The method is shown to be effective in finding miRNA targets. The method is based on 18O/16O labeling that can be used in many applications. Aimed to reduce false positives to a minimum, while accepting loss of probable targets of miRNAs	Ma et al., (2015)
To perform metabolomics by pseudo-targeted approach to biological samples	A software developed to automatically select MRM ion pairs and optimal CE voltages	Fasting serum from healthy and hepatocellular carcinoma (n = 30).	MRM-Ion Pair Finder and Marker View	Devised software to automatically select most intensive ion product, and quantify it better. It combines the advantages of scanning of metabolome by untargeted approach, while maintaining the accurate quantification achieved by multiple reaction monitoring.	Luo et al., (2015)

N/A, not applicable.

Another tool that pointed out errors in the annotation of dominant miRNA strands using an advanced sequencing method is AQ-Seq (Kim et al., 2019). It overcomes the bias in conventional methods and helps to identify miRNAs with correct 5’ end, their isomiRs, and modifications, all of which are important in target prediction and some of them can predominantly distinguish tumor from normal tissues across 32 TCGA cancer types (Loher et al., 2014; Telonis et al., 2015, 2017). This can probably explain the observed discrepancies between miRNA and target levels in real experiments due to differential targeting by isomiRs and help to quantify miRNAs and isomiRs accurately, and better understand how these modifications affect their targets when combined with other omics.

A recent study highlighted a method to identify spatially resolved and multiplexed miRNA tissue array to determine the spatial distribution of miRNAs in tumor tissues (Ho and Lee, 2015; Nagarajan et al., 2020; Turunen et al., 2019). The array has wells with reagents that can release miRNAs from the tissue, and then probes polymerized in hydrogel can quantify the released miRNAs. This can open new avenues to visualize tumors better and predict disease prognosis and treatment response in complex samples (Nagarajan et al., 2020).

To overcome the difficulties with commonly used MS-based proteomics methods such as extensive labeling protocol, high costs, and limitations in detection of complete proteome, a novel method was developed utilizing a new O16/O18 labeling method combined with other approaches and statistical tests to identify truly significant differentially expressed proteins and recognize the strongly positive targets of miRNAs of interest from them (Ma et al., 2015).

In continuation with the above developments, a novel tool multiple reaction monitoring (MRM)-ion pair finder was developed to combine the advantages of targeted and untargeted approaches for protein selection and quantification and sufficiently reduce time on manual selection of appropriate ions for quantitation as in targeted method. It helps to identify both known and unknown metabolites from real biological samples with significant accuracy and reliability and reduced false positives (Luo et al., 2015).

The above developments allow an indirect quantification of miRNAs without the hassles of amplification and use of fluorophores and can be expanded for a panel of miRNAs to be used as prognostic or diagnostic markers in a clinical setting. Several such indirect quantifications developed earlier can also be used to detect and/or quantify a single miRNA or a panel of miRNAs by amplification or a unique probe readout, which have possible limitations as qRT-PCR, warranting further processing and refinement (Cadoni et al., 2020; Cheng et al., 2018; Dave et al., 2019).

Overall, these improved and advanced omics technologies can increase the reliability of data obtained and if integrated together with other omics can provide a significant knowledge base for miRNA interactome and its undiscovered characteristics. They can be applied in the final step of multiomics study for a given phenotype or in intermediate steps to generate an inclusive list of biomarkers.

Multiomics as a Tool to Study miRNA Interactome

Multiomics approach includes a combination of single-omics technologies such as next-generation sequencing and MS for genomics, epigenomics, transcriptomics, proteomics, and metabolomics (Aebersold and Mann, 2003; Dettmer et al., 2007; Kulski, 2015), which gives a unique set of data with its advantages and disadvantages. The wide range of molecules interacting with a miRNA makes it evident that the integration of multiple data types is necessary to obtain a holistic picture of miRNA interactome. The significance of multiomics in providing a better understanding of diseases by extracting the most from a single layer of data of different types and analyses of various strategies to understand disease-causing agent has been studied (Hasin et al., 2017).

Use of different omics approaches combined with bioinformatics for predictive, preventive, and personalized medicines (PPPMs) has been reported in cancers (Yoo et al., 2018). Combined or integrated omics, such as proteogenomics and proteometabolomics, is currently gaining momentum in various cancers to identify biomarkers for PPPM and can accelerate the detection and prevention of cancer. Different algorithms are used in integrating multiomics to give meaningful information that might help in clinical applications (Huang et al., 2017).

However, these methods include one or a few omics, particularly gene expression, miRNA expression, copy number variations (CNV), and methylation status and miss out on the other omics. In addition, there is a lack of methods to integrate data from multiple cohorts, which is the demand of the hour to generate useful clinical data. Employing large number of subjects for analysis to gain meaningful insight from multiomics data is a limitation, but the increasing availability of large-scale human studies like Metabolic Syndrome in Man (METSIM) (Laakso et al., 2017) and Genotype Tissue Expression (GTEx) (Aguet et al., 2017) mitigates this. In this direction, Chakraborty et al. (2018) reviewed those reports involving a minimum of two or more omics, mostly the cancer repository: the cancer genome atlas (TCGA) (Weinstein et al., 2013) data.

Several interesting observations came out of multiomics, such as a) diversified applications of radiomics, a bridge between imaging and personalized medicine, with other omics and b) the use of system biology to develop strategies for personalized medicine (Lu and Zhan, 2018; Yoo et al., 2018). A common observation (Chakraborty et al., 2018; Yoo et al., 2018) is that single omics cannot indicate the causative relationship between the observed signatures and the disease, highlighting the advantage of multiomics over single omics.

Taken together, these reports indicate a significant understanding of multiomics in cancer prevention, treatment, and prediction of disease progression with a lack of discussion on miRNA interactome involving multiomics in cancer. Thus, in this review, we attempted to bridge this gap by focusing on two key aspects: (1)

Role of multiomics in studying miRNA interacting partners with various approaches and bioinformatics tools

(2)

The future applications of multiomics in miRNA and conventional therapy combination approach and treatment response optimization.

Figure 3 illustrates various outcomes of integration of multiomics for miRNA interactome characterization. An interactome network between miRNAs and targets can showcase molecular interactions in the form of nodes (units of the network) and critical hubs (nodes with the greater number of interactions than average) with diagnostic or prognostic potential.

FIG. 3.

Multiomics in miRNA interactome and its possible outcomes.

Application of Multiomics in miRNA Interactome Identification

Two different approaches can be used to understand how a miRNA or multiple miRNAs work together during the progression of cancer and if they can act as prognostic/diagnostic markers and/or therapeutic targets, (1) a disease model with its phenotype and a pattern involving a miRNA or multiple miRNAs and (2) the effect of a miRNA or its cluster on its interactome.

A Disease Model-Based Interactome Identification

In this approach, the first step is to characterize different phenotypes modulated by miRNAs such as metastasis, downregulation of specific protein, therapy sensitivity, and others. A list of the studies discussing the disease model-based interactome identification, including omics and software used for analyses, is shown in Table 2.

Table 2.

Various Omics Platforms and Software Used for Multiomics Analyses

Multiomics	Methods	MicroRNA	Molecules involved	Prediction tools/softwares	Outcome	Reference
MiRNA profiling and proteomics after DDX5 knockdown	SILAC-based proteomics; miRNA profiling using miRCURYTM LNA Array (fifth version, Exiqon)	miR-182 and miR-21 by DDX5 knockdown	DDX5 (RNA helicase) involved in miRNA processing and maturation	TargetScan, DAVID, STRING	DDX5 serves as a master regulator for a subset of miRNAs, leading to actin cytoskeletal changes	Wang and Huan (2012)
MiRNA expression profiling and proteomics after cigarette smoke treatment	Microarray-based miRNA expression profiling, and stable isotopic labelling with amino acids in cell culture-based quantitative proteomics analysis, TCGA miRNA expression data for comparison	miR-124, miR-21, miR-192, miR-125, miR-374a, miR-204, miR-222, miR-145, miR-184, miR-155, miR-196a	No targets discussed in particular; miR-222 targets-PPP2R2A and HMGA1; miR-204 targets-MMP1 and SOD2; miR-145 target-EGFR; miR-184 target-PDCD4; miR-155 target-CEBPB; miR-196a target-S100A9.	Target Scan, FunRich, miRTarBase	Identified an altered expression of several miRNAs as well as proteins in response to cigarette smoke. Pathway analysis This study brings together all the markers under the integrated approach, validating the study.	Advani et al., (2017)
Proteomics profile; targets validation in lung adenocarcinoma and H₂O₂-treated samples; miRNA microarray to identify miRNAs targeting PFKm (phosphofructokinase mutase, muscle)	Proteomics by 2D-DIGE combined with (MALDI-TOF) MS and TOF/TOF tandem MS; miRNA microarray	miR-320a	ETS-1 regulating miR-320a, which in turn regulates PFKM	Targetscan	Reported 1. Differentially expressed proteins in MV 2. Genes in lung cancer. 3. miRNA targeting select proteomics targets in MV	Tang et al., (2012)
363 hepatocellular carcinoma (HCC) cases analyzed by whole-exome sequencing and DNA copy number analyses, and a core of 196 HCC cases analyzed by DNA methylation, RNA, miRNA, and proteomics expression.	Illumina sequencing, reverse phase protein array analysis, Illumina Infinium DNA methylation platform HumanMethylation450 (HM450), Affymetrix SNP 6.0 arrays	miR-122	Somatic mutations: two, LZTR1 and EEF1A1, SF3B1 and SMARCA4 displayed mutations identical to those identified as driver mutations in other cancers; Significantly mutated genes-APOB and ALB downregulated; WNT or p53 signaling or the telomerase promoter are altered in 77% of HCC patients; RTK inhibitors as therapeutic strategy; immune microenvironment changes activating effector cells to resting suppressive immune cells	MutSigCV algorithm, Bayesian variant of the non-negative matrix factorization (NMF) algorithm, GISTIC 2.0., multivariate regression approach	Understanding of molecular targets leading to improved therapeutic strategies by integrated analytic approaches using multiple data platforms from a large set of clinically annotated HCC	Wheeler and Roberts (2017)
MiRNA profiling between nonmetastatic and metastatic cell lines; identified upregulated miR overexpression in nonmetastatic cell line and compared by proteomics with metastatic cell line	TaqMan array microRNA cards for global miRNA profiling; iTRAQ proteomics of upregulated miRNAs	MiR-424-3p, miR-503, or miR-1292	41 miRNAs commonly dysregulated across all 3 miRNA-overexpressing cell lines	miRwalk, Segal Lab miRNA	A panel of at least three miRNAs and two target proteins was identified, which constitute relevant candidates for therapeutic intervention in liver metastasis.	Torres et al., (2018)
Reverse protein array of LNCaP cells with mimetics corresponding to 12 SiM (suppressed in metastasis) miRNAs (miR-143-5p and miR-204 were excluded) vs. control, compared newly obtained data with a multitude of previously obtained data, for example, HITS-CLIP and PAR-CLIP data, transcriptomics data of previous si-RNAs for AR silencing and metastatic prostate cancer (PC) samples	Proteomics using RPPA by a panel of 157 antibodies (41 against phosphorylated epitopes, 4 against cleaved epitopes, and 112 against total proteins)	miR-135a-5p	Pathways involved in metPC are cell cycle, apoptosis, AKT/mTOR signaling and the AR axis, AR signaling and co-receptors	Consensus Path DB source for pathway analysis, protein expression was quantified by MicroVigene and normalized using the R package SuperCurve, TargetScan, PicTar, miRanda, mirDB, and DianaLab	Identified 1. SiM miRNAs from different GEO datasets, 2. Methylation status by Illumina Infinium Human Methylation 450 K BeadChip Kit, 3. Prognostic significance of miR-MTT assay, 4. Pathways affected by SiM-MiRs to narrow down on interesting miR-135a-5p/AR axis.	Coarfa et al., (2015)
Matched miRNA and mRNA expression data from nine CRC cell lines were downloaded from the Gene Expression Omnibus (GSE10833 and GSE10843), 9 cell lines (DLD-1, HCT-15, COLO 205 and SW480, HCT-116, HT-29, RKO, Caco-2, and LoVo.	MS—Liquid chromatography-tandem MS (LC-MS/MS) shotgun proteomics analyses, Matched miRNA and mRNA expression data from the same nine CRC cell lines were downloaded from the Gene Expression Omnibus (GEO, GSE10833, and GSE10843).	miR-138,miR-141, miR-200c	No targets discussed in particular	TargetScan, miRanda, or MirTarget2	Indicates that perfect matches to both the seed and 3’ portion of miRNA favor mRNA decay more than translational repression, whereas structural subtleties of imperfect miRNA-mRNA duplexes may preferentially trigger translational repression.	Liu et al., (2013)
Comparative miRnome analyses of MCF-7 [ER-β +ve (c-ter and N-ter Tagged vs ER-β -ve)] and Zr-75 cells [ER-β +ve (FLAG Tagged) vs ER-β –ve] quantitative proteomics of MCF-7 cells stable cells expressing c-ter-ER-β vs wt MCF-7, MCF-7 cells transfected with mimic miR-30a vs sham oligo transcriptomics and proteomics	Massively parallel next-generation miRNA sequencing, gene expression profiling, and isobaric tags for relative and absolute quantitation (iTRAQ) quantitative proteomics	miR-30a-5p; upregulated in both cell lines and in both nuclear and cytoplasmic compartments of MCF-7	EPB41L5, TPD52L1among 11 other proteins overexpressed in human breast carcinomas	iMiR, Deseq, ParagonTM search algorithm using the ProteinPilot 3.0 interface, GenomeStudio software, miRanda, miRDB, TarBase, and TargetScan	ER-β +ve vs ER-β-ve profiling identified miRNAs dysregulated in ER-β+ and proteins dysregulated in ER-β+ve compared to ER-β-ve; identified miR-30a-5p as the most upregulated miRNA; found proteome and mRNA changes of the same. This snapshot of proteins modulated by miR-30a-5p also identified by proteomics in unliganded ER-β+ve highlights that miRNAs are key players in modulating the response of cells to ER-β in the absence of any ligand.	Nassa et al., (2014)
MiRNA and protein expression profiles were obtained from 71 estrogen receptor-positive (ER-β) and 25 triple-negative breast cancer (TNBC) samples, RNA and protein from FFPE (Formalin-fixed paraffin-embedded) tissues	LC-MS/MS analysis; TaqMan Array MicroRNA Card using miRNA profiling; orthogonal comparison done using gene expression datasets (1296 breast carcinoma) to that of proteins identified in proteomics.	miR-301a, miR-331-5p, miR-301a, miR-331-5p, miR-135b, miR-190b, miR-224, miR-449a, miR-23b, miR-375	Several proteins converged into separate groups as identified by Probabilistic graphical model	MaxQuant followed by protein identification using the integrated Andromeda search engine, significant analysis of microarrays (SAM), miRWalk, DAVID analysis	Measured the expression of 1,095 proteins and 90 miRNAs in breast cancer samples, unraveling differences in the activity of molecular processes, such as proliferation and metabolism, between ER-β and TNBC tumors, which were also confirmed at the transcriptome level. This gives an overall picture of changes between 2 molecular subtypes and is a valid approach.	Berges-soria et al., (2015)
Microarray and NGS to identify miRNAs in EGFR-driven pathways in MDA-MB-231, RPPA for EGFR pathway-related proteins (previously published RNA-seq data used to identify the genes of the pathway with at least one transcript in chosen model, one available validated antibody for RPPA and only these genes used in RPPA screen)	RPPA to mimic transfected cell lysates (genome-wide) for 26 proteins of EGFR pathway only in MDA-MB-231 cell lines and validated selected miRNAs in MCF-7 and U87 cells	810 different miRNAs, validated-miR-124, miR-147, miR-193a-3p	Cell cycle proteins such as RB1, PTEN, CDKN1B, CDK4,CCND1 etc.	MicroRNA.org for identifying miRNA-mRNA seed matches, TargetScan to identify evolutionary conserved sequence matching filter, TargetScan and microCosm for comparing predicted vs nonpredicted miRNA for 26 proteins	Identified miRNAs that regulate the EGFR-driven cell cycle pathway. Scaled up the tested miRNAs to the whole-genome level, while scaling down the number of proteins to a single pathway. Several cell cycle components that regulate the G1/S transition check point were found to be co-regulated.	Uhlmann et al., (2012)
RPPA done against myc, pRb 1, and CDKN1B for 879 miRs in MDA-MB-231 cells; a second screen against these proteins were done using a subset of 56 miRs from the first screen and another set based on multiple criteria, including downregulation of previously validated targets and regulation of previously invalidated targets of interest.	RPPA done against myc, pRb 1, and CDKN1B for 879 miRs in MDA-MB-231 cells transfected for 48 h	miR-124 leading miRNA identified along with 4 others (miR-18a, miR-34b, miR-365, and miR-506)	ITGB1 (decreased) and CDKN1B (increased) by miR-124	TCGA data from cBioportal	Identified proteomics signature in vitro study to screen miRNAs and used clinical data (different cancer types) was used to filter and ascertain the leads. miR-124 was validated for tumor-suppressive functions both in vitro and in vivo (nanoparticle delivery)	Seviour et al., (2016)

When DDX5 (DEAD-Box Helicase 5), an RNA helicase and noncoding RNA processing enzyme upregulated in breast cancer, was knocked down, 2 proteins (profilin and cofilin) were found to be upregulated (by proteomics) (Wang and Huan, 2012). Furthermore, miRNA profiling revealed DDX5 downregulation by a subset of miRNAs (miR-21, let-7 family, and miR-182), bioinformatics analysis of which revealed miRNA-182 to be responsible for the observed phenotype, providing a unique target for therapy against basal-type breast cancers.

In another study, the comparison of differentially expressed miRNAs from cell line studies (chronic exposure to cigarette smoke vs. parental) with TCGA datasets (smoker vs. nonsmoker) revealed overlapping miRNAs, confirming the in vitro study, following which combined miRNA profiling and proteomics analysis identified miRNA target pairs and pathways dysregulated in response to cigarette smoking, giving an integrated picture of the effect of cigarette smoking on cells (Advani et al., 2017).

Another study identified that glycolysis is commonly dysregulated in lung cancer and in disused diaphragm muscle by proteomics and used microarray to identify miRNA(s) regulating the rate-limiting step (Phosphofructokinase) in glycolysis, elucidating the mechanism of these disease phenotypes (Tang et al., 2012).

The combination of five omics data namely DNA copy number, DNA methylation, mRNA expression, miRNA expression, and reverse phase protein array (RPPA) was used to identify unique signatures in hepatocellular carcinoma (HCC) by TCGA group, to classify the subtypes in HCC: iclust 1, 2, and 3, with iclust 1 being the one with lower survival. This study has also identified different gene signatures that can be used to provide targeted therapy (Wheeler and Roberts, 2017). Furthermore, in this study, a minor class of HCC (iclust 1 subclass with isocitrate dehydrogenases (IDH) mutations), identified through integrative analyses of mutation and gene expression followed by miRNA expression, revealed miR-122 as a dysregulated miRNA for IDH mutant and IDH-like samples with promoter hypermethylation as the mechanism of dysregulation with its targets identified by miRNA-mRNA anticorrelation studies.

Together, this study has classified HCC subtypes, identified a minor subtype with poor prognosis, and a prospective target (miR-122) for the same subtype. This study highlights the advantages of multiomics studies emphasizing that the research integration from different reports can progressively develop therapeutic targets precisely than independent investigations.

In another investigation in colorectal cancer, identification of differentially expressed miRNAs has led to a group of 3 upregulated miRNAs (miR-424-3p, miR−503, and miR −1292) associated with overall survival and disease-free survival (Torres et al., 2018). These miRNAs share common targets identified by proteomics studies, while none of the bioinformatics tools predicted even one of them. This unique approach combining miRNA transcriptomics and proteomics gives a set of miRNA-target pairs that can be used as therapeutic targets to stall or prevent metastasis.

Another study in colorectal cancer elucidated the mechanisms of miRNA-mediated effects by combining miRNA and mRNA expression profiling with proteomics data (Liu et al., 2013) and revealed different mechanisms (mRNA Decay; mRNA Decay with other mechanisms; translational repression; translational repression with other mechanisms; mRNA decay; and translational repression, both strong or both weak) for miRNA regulation. Also, it highlighted the features (8mer site, site positioning within 3'UTR, local AU-rich context, and additional 3’ pairing) responsible for the same, which could be potentially used to develop newer computational target prediction tools.

Nassa et al. (2014) attempted to reveal the mechanism of a relatively less studied estrogen receptor β (ESR2) in breast cancer, which mediates estrogen-independent effects (Nassa et al., 2014). Although many dysregulated proteins were identified as targets for multiple miRNAs with miRNA profiling and proteomics profiling, the correlation between mRNA and miRNA was only moderate, indicating translational repression as a major mechanism. A subset validation of a single differentially expressed miRNA (miR-30a-5p) and its target mRNA (SNAI1), and protein levels found it to be correct. Molecular changes deduced between estrogen receptor-positive (ER+) and triple-negative (TNBC) breast cancer subtypes by utilizing differential miRNA and proteomics profiling and system biology approach revealed a differential single functional module of miRNAs and proteins (Berges-soria et al., 2015). This multiomics strategy involving system biology paves the way for envisaging the system in action and reveals novel therapeutic targets.

Reverse phase protein array (RPPA) is a low-throughput proteomics method based on antibody validation, but is still used when a specific phenotype is being addressed; an miRNA signature (miR-1, miR-133a, miR-133b, miR-135a, miR-143-3p, miR-145-3p, miR-205, miR-221-3p, miR-221-5p, miR-222-3p, miR-24-1-5p, and miR-31) was found to be suppressed in metastasis in multiple datasets, with relevance to prostate cancer (PC) (Coarfa et al., 2015).

Mechanisms of this dysregulation were identified by DNA methylation, copy number alteration signatures, and prognostic value through previously available datasets. Proteomics analysis of these miRNAs with RPPA revealed only a fraction of interactions of miRNAs with targets to be direct, reiterating the importance of understanding both direct and indirect effects of miRNAs as stated earlier.

Furthermore, Uhlmann et al. (2012) tried to overcome the limitations of RPPA such as antibody inaccuracy and utilized proteomics analysis to identify miRNAs regulating a single pathway. This led to the development of a low stringency network consisting of those targets bearing at least one 6mer evolutionary conserved site in 3’ UTR, showcasing the interactions of co-regulated proteins by one or more miRNAs (miR-124, miR-147, and miR-193a-3p) in pathway modulation.

An end to end study (Seviour et al., 2016) used a powerful iteration of clinical (TCGA) and in vitro data to determine the clinically relevant therapeutic targets for an RPPA-identified signature (CDKN1B, Myc, and phospho-Rb 1) in breast and ovarian cancers. Then, five corresponding miRNAs (miR-124, miR-365, miR-34b*, miR-18a, and miR-506) targeting this signature were identified from the proteomics data (TCGA and RPPA data of cell line models). The miRNA list was further refined to a single miRNA (miR-124) by validation on the other cancer models and strengthened by evidence in drug sensitization using the in vivo models. This report thus provides a holistic picture, resulting in a robust therapeutic target utilizing a plethora of data and omics.

Effects of miRNA and Its Clusters on Its Interactome

Another approach that uses miRNA-based proteomics to identify its interacting partners and mechanism of action has been reported (Giambruno et al., 2018) and several other reports where miRNA targets/interactome identified by proteomics are shown in Table 3.

Table 3.

MiRNA-Based Interactomes

MicroRNA	Model	Multiomics (proteomics) method(S)	Software used for proteomics analysis	Multiomics (nonproteomics) methods	Software used for multiomics analyses	Outcome	Reference
MiRNA-34a	Neuroblastoma	Proteomics using isotope-coded affinity tags (ICAT)	Turbo SEQUEST against human protein database from EBI to identify peptides, XPRESS to measure relative abundance of peptides	mRNA profiling by Affymetrix U133plus2 microarray	NA	Ribosomal proteins are the only functional protein class significantly enriched in the group of miR-34a-downregulated proteins and is also regulated by YY1. Limitations - Not many experimentally validated targets were noted by this method.	Chen et al., (2011)
MiRNA-143/145	Colon cancer	LC−ESI−MS/MS coupled to a LTQ Orbitrap Velos mass spectrometer, SILAC-based proteomics	Maxquant software, Skyline v1.0	Affymetrix Gene Chip Human Exon ST 1.0 arrays (Affymetrix, Santa Clara, CA)	Bioconductor software package with mean fluorescence intensities derived from a log₂transformation of the data and normalization using the quantile normalization method.	Combined analyses highlighted the differential effects of miR-143/145 cluster versus individual miRNAs indicating a secondary regulation of targets by the clusters	Bauer and Hummon (2012)
MiRNA-200 family	Breast cancer	RPPA for cell cycle proteins and ZEB-1 for miR-200 family members, mRNA microRNA profiling to identify targets of miR-200bc/429	Signal intensities were quantified using Gene Pix Pro 5.0 (Molecular Device, Sunnyvale, CA, USA)	Sentrix HumanWG-6 arrays (Illumina, San Diego, CA, USA) were used for mRNA profiling.	Expression profiling data were analyzed by Bioconductor limma package	Differential effects of miR-200 family members on invasion, cell viability, cell proliferation, apoptosis, and cell cycle were studied and identified miR-200bc/429 as the dominant regulator of these observed phenotypes. Both clusters can act as tumor suppressors.	Uhlmann et al., (2010)
Let 7 family	Osteosarcoma	Nano HPLC, electrospray ionization quadrupole time-of-flight mass spectrometer, quantitative shotgun proteomics analysis	2DICAL, Mascot	3D-Gene TM Human microRNA Oligo chips (Toray, Tokyo); 939 probes represented 866 known human miRNA and 73 virus-originating miRNA	N/A	Let-7 family and its targets were identified uniquely to promote osteosarcoma	Kobayashi et al., (2014)
MiRNA-145	Pancreatic cancer	SILAC-based proteomics LC-MS/MS analysis carried using LTQ-Orbitrap Velos mass spectrometer	Mascot and SEQUEST, Proteome Discoverer platform	mRNA microarray on Agilent-014850 Whole Human Genome Microarray platform; miRNA microarray on Agilent's-Human miRNA Microarray platform (V3).	N/A	Identified miR-145-mediated pathway regulation through both proteomics and transcriptomics	Huang et al., (2015)
MiRNA-143	Pancreatic cancer	SILAC-based proteomics LC-MS/MS analysis carried using LTQ-Orbitrap XL ETD	ProteomeDiscoverer (Ver 1.1) software workflow, which combines data from Mascot and SEQUEST search engines and performs the quantitation; Quantitation by MSQuant is done by calculating the ratios of heavy and light peptide intensities measured at different time points of LC peaks and merged together with ProteomeDiscoverer	Agilent gene expression arrays, which contain 41 000 unique probes for transcripts, including 26,705 genes	GeneSpring GX 10.0.1	Identified miR-143 through both proteomics and transcriptomics	Yang et al., (2010)

N/A, not applicable.

A combination of proteomics and mRNA microarray was used to identify miR-34a targets (Chen et al., 2011) using isotope-coded affinity tags (ICAT) labeling and reverse phase liquid chromatography (RPLC)-MS/MS to identify novel targets with higher efficiency compared to matrix-assisted laser desorption ionization time of flight (MALDI TOF) MS/MS (Cheng et al., 2010). This shows that among proteomics techniques, each has its advantages and disadvantages, and therefore multiomics is a necessary tool to overcome the drawbacks of single proteomics techniques (Li et al., 2012).

An earlier study combined the proteomics and microarray analysis to show differential targets between an individual miRNA (miR-143 or miR-145) and its cluster (Bauer and Hummon, 2012). A similar study combined mRNA microarray and RPPA with prediction programs to identify direct targets of miR-200bc/429 to understand how the differential effects of 2 clusters of miRNA-200 family are mediated through their exclusive targets (Uhlmann et al., 2010). Although this is not a large-scale multiomics study, it succeeded in understanding the two different miRNA clusters of the same family differentially affecting the phenotype in the same cell line.

A miRNA signature from clinical samples followed with proteomics study of a single family (let-7 family) to identify the common targets of the family was reported (Kobayashi et al., 2014), similar to the strategy of disease model-based miRNA interactome identification. Unlike the Kobayashi study, two separate studies identified targets of miRNA-143 and miRNA-145 using proteomics, and they compared mRNA and miRNA changes to understand the mechanism of targeting by miRNA-143/145, that is, either mRNA degradation or translational repression (Huang et al., 2015; Yang et al., 2010).

These studies used a single miRNA proteomics approach that helped identify targets of single miRNA/family, but unlike the first disease model-based approach, it does not present us with a holistic view. However, this approach helps in enlisting the unique properties of miRNA regulation and its targets on a large scale. Most of the later studies focused on the disease instead of a single miRNA, to understand multiple dimensions of the disease.

Role of Multiomics and miRNA Interactome in Therapeutic Efficacy

Novel strategies such as combination therapy of tumor suppressor miRNAs with conventional therapy appear promising to overcome the resistance to conventional cancer therapies and prevent the recurrence of the disease (Ebrahimi and Hashemy, 2019; Mognato and Celotti, 2015). Studies with individual miRNAs predict sensitivity to therapy, while multiomics studies identify biomarkers that correlate with drug sensitivity or establish the mechanism of action of novel compounds.

A large-scale multiomics study utilized a combination of different datasets (mentioned in Table 4) and gave a pretty robust signature of genes, miRNAs, mutations, copy number alterations, and transcription factors, that is, biomarkers to predict responders and nonresponders (Bosquet et al., 2014). Another omics study of a combination of mRNA, miRNA microarray, and proteomics was utilized to understand the mechanism of action of a novel compound (shikonin) against leukemia (Wiench et al., 2013). Furthermore exploratory investigations on these studies will help identify those strongly modulating miRNAs that can be used as radiosensitizers or chemosensitizers. Immunotherapy is also being used in cancer treatment involving two major strategies: checkpoint inhibitors and adoptive cell therapy (tested in multiple cancer types like glioblastoma) (Matsiko, 2018).

Table 4.

MiRNA-Based Multiomics Data Applications to Predict Therapeutic Response

Model	Therapy resistance	MicroRNA	Multiomics	Outcome	Reference
Ovarian cancer	Chemotherapy (platinum based)	38 miRNAs differentially expressed between CR and IR (i.e., complete response (CR) to therapy vs incomplete response (IR) to therapy)	Copy number alterations, gene mutations, DNA methylation, miRNA expression, and transcription Factor binding site location	The cumulative analysis of all these datasets helped identify a robust signature of 422 genes correlated to pathways involved in chemotherapy response	Bosquet et al., (2014)
Leukemia	Chemotherapy, identification of new compound shikonin to overcome incurable leukemia patients by chemotherapy	150 microRNAs associated with predicted and experimentally observed targets	mRNA, miRNA microarray, and SILAC proteomics	The combined analysis identified PI3K-Akt-mTOR as primary target of shikonin, mitochondrial dysfunction, as well as overcame resistance developed by monotherapy	Wiench et al., (2013)
Glioblastoma	Immunotherapy	miRNA-216a,b miRNA-708 and let-7i	miRNOmics by Illumina-seq and proteomics (TMT10plex-based isobaric quantification approach)	Combined analysis of proteomics and miRNA analysis of dendritic cell vaccinated patient samples resulted in a consensus pathway that showed association with overall survival and also presented a novel idea of sensitizers for immunotherapy from clinical data	Erhart et al., (2020)
Pan-cancer	Chemotherapy	No microRNA in particular	RPPA-based proteomics data from TCGA; CNV, gene and miRNA expression, mutation, and MS-based proteomics data from previous datasets for NCI-60 panel/DREAM7 panel	Multiomics resulted in improved drug response predictive power; MS-based proteomics was slightly better at prediction than RPPA-based proteomics; however, RPPA-based proteomics became important to predict the drug response of molecular targeted therapies	Ali et al., (2018)

In a first of its kind multiomics study to hypothesize sensitization to immunotherapy, proteomics and miRNA microarray were used to identify the survival-associated biomarkers to determine response to tumor lysate-charged dendritic cell (DC) vaccination (Erhart et al., 2020). There is a dearth of literature in multiomics with miRNAs for therapy resistance, although many individual studies where a single miRNA or miRNA family such as miR-34 have been implicated in radiation response as well as their role in radiosensitizing cells (Chaudhry, 2013; Marta et al., 2015; Moertl et al., 2016).

Only one study showcases the predictive power of drug sensitivity in cancer cells by single-omics technologies, that is, proteomics methods, MS and RPPA, revealing that one could be better than other technologies for prediction of sensitivity, but multiomics components can complement each other and overcome shortcomings, while adding new information to the pool, drastically improving the predictive power (Ali et al., 2018). Multiomics can play a key role to understand the global response of cells to cytotoxic agents (radiation, drugs, or immune cells), to identify key signature molecules/targets/miRNAs in resistant tumors and in personalized treatment plans, to enhance response in resistant tumors. A summary of miRNA-based multiomics for therapy response as well as key cancer response pathways is depicted in Table 4.

Bioinformatics Tools to Predict miRNA Interactome from Proteomics Data

A list of tools (input/output data, programming language used, model used in development) that are available to identify miRNA targets by incorporating proteomics data as well as to identify its interactome is given in Table 5 (Li et al., 2009; Qin et al., 2013; Reczko et al., 2012). Some of the models used in these tools predict targets from input data either by probabilistic determination or machine learning. There is a lack of interest in developing applications for omics integration in miRNA target prediction, although there are several independent tools for analyzing quantitative proteomics data and miRNA target prediction. Since miRNAs exhibit unique regulation, newer methods are required to correctly dissect direct and indirect miRNA interactions with targets identified from proteomics.

Table 5.

Bioinformatics Tools to Forecast MiRNA Interactome from Proteomics

Objective(S)	Assumptions	Prediction programs	Input data	Output data	Model	Tool/algorithm	References
Probabilistic model to forecast by combining sequence complementarity, miRNA expression level and protein abundance	Assumption 1: the miRNA expression is high and the protein abundance of the mRNA is low Assumption 2: Trace back to the mRNA expression of the confident targets to investigate the mechanisms of the miRNA-mediated post-transcriptional regulation	TargetScan, miRanda or PicTar	Putative miRNA targets from target prediction programs, protein abundance data from mass spec data, gene expression data from microarray studies	Scores miRNA-mRNA pairs and identifies the mechanism	Bayesian	None	Li et al., (2009)
The ANN is trained to reproduce experimentally measured protein repression rates and thus its final output better reflects the strength of the miRNA:target gene (miTG) interaction	N/A	PicTar, TargetScanS, TargetScan5.0, PITA, ElMMo, RNA22, miRanda, miRBase-Targets	High throughput proteomics data of miRNA overexpression or downregulation	Uses high-throughput proteomics experiments to optimize the separation of miTGs and non-miTGs and achieving highest performance	Machine learning ANN	DIANA-microT-ANN	Reczko et al., (2012)
Developed to address miRNA-centered regulatory networks from high-throughput proteomics data	N/A	TargetScan, PicTar, miRanda, PITA, experimental databases include starBase, miRecords, and TarBase	Inputs in tab-delimited format containing high-throughput protein and mRNA expressions	The inferred miRNA-centered regulatory network is visualized using Cytoscape	N/A	ProteoMirExpress	Qin et al., (2013)

N/A, not applicable.

Future of Multiomics

Currently, most of the multiomics studies aim to identify prognostic and diagnostic markers (Kong et al., 2020; Kwon et al., 2015; Yang et al., 2019) and classify subgroups of different cancers (Aure et al., 2016; Zhang et al., 2016b), driver mutations in subgroups of tumors (Castro-Vega et al., 2015), markers to track the relapse of treated patients (Madhavan et al., 2013), and metastasis (Kong et al., 2020), database generation in cell lines (Berg et al., 2017) as well as for use in precision medicine (Xie et al., 2018).

It will be of interest to also incorporate information on sequence variants in the miRNA and its interacting partners as a part of the global interactome identification, thereby serving the needs of precision and personalized medicine in a clinical setting. Development of miRNA regulatory pathways in a pan-cancer model and domestic animals will help pinpoint common/synergistic regulatory pathways and generate novel therapies targeting/modulating miRNAs on different human and zoonotic diseases (Pasquini and Kunej, 2019; Zhang et al., 2016a).

Attempts are also being made to generate markers from pan-cancer datasets by integrated clustering methods and to compare their effectiveness as well as similarities between the various methods (Wei et al., 2020). It will become imperative to detect cancers early using markers from multiomics data or to determine novel markers or patient outcome, which could be pan-cancer in nature (Mitra et al., 2020). It will also be advantageous to incorporate live patient data in real time using automated methods for integration and analysis of multiomics data, all of which can be a big step in accessing real-time data and apply the knowledge in personalized medicine. Tools that perform end-to-end analysis in a single package format are also needed, and attempts are being made in that direction, utilizing multiomics for wider applications (Allendes Osorio et al., 2020).

Currently, advancements in multiomics focus on the integration of data from different omics and novel algorithms/software are being developed, despite the various issues (data handling, annotation, design, assumptions, and statistical analysis) in data integration (Misra et al., 2019). Integration tools of multiomics operating on different principles (Sathyanarayanan et al., 2019) as well as the tools for data visualization and data repositories are under development (Subramanian et al., 2020), and several of these tools incorporate miRNA expression data in the integration of omics data. Some of the recently published tools, especially with respect to miRNA and/or proteomics, with their advantages, input omics datasets, and final output have been summarized in Table 6.

Table 6.

Contemporary Tools for MultiOmics Integration Including Either Proteomics or miRNA Sequencing Data

Objective(S)	Tool name	Omics dataset	Output	Program and availability	Advantages	Reference
Integration and analysis of multiomics data	Miodin	Horizontal data integration from different studies of differentially expressed RNA in lung cancer, transcriptomics data Vertical data integration from RNA-seq, methylation data and microarray from breast can,cer samples in TCGA	Horizontal and vertical data integration leading to multiomics data analysis with reproducible results	R package	A clean workflow, ensures reproducibility by uploading the docker image on a folder, which can be accessed by others and verified	Ulfenborg (2019)
Network-based integration of multiomics data and Wasserstein distance clustering	aWCluster	mRNA expression, DNA copy number alteration, and DNA methylation, as well as the interaction network connectivity of gene products from invasive breast cancer TCGA	It can cluster TCGA data to classes with different survival rates	MATLAB	Combination of multiple regulatory levels in cancer biology and gene (protein) interaction network with mathematical models for predictive treatment plan development and personalization of cancer patient treatment	Pouryahya et al., (2020)
Multiomics clustering on partial data	NEMO	Gene expression, methylation, and miRNA expression on different cancer datasets Acute Myeloid Leukemia (AML), Breast Invasive Carcinoma (BIC), Colon Adenocarcinoma (COAD), Glioblastoma Multiforme (GBM), Kidney Renal Clear Cell Carcinoma (KIRC), Liver Hepatocellular Carcinoma (LIHC), Lung Squamous Cell Carcinoma (LUSC), Skim Cutaneous Melanoma (SKCM), Ovarian serous cystadenocarcinoma (OV) and Sarcoma (SARC)	NEMO can be utilized to cluster partial datasets of different omics data with similar efficiency in full datasets of different omics data	https://github.com/Shamir-Lab/NEMO	Fast and much simpler than existing multiomics clustering algorithms, and avoids iterative optimization	Rappoport et al., (2019)
Multiomics-unsupervised clustering for subtype identification	PINSPLUS	Two METABRIC breast cancer datasets and 34 TCGA datasets	It outperformed other tools (CC, SNF, and iClusterPlus) with respect to survival analysis per subtype	R program	Fast and powerful software for subtype discovery	Nguyen et al., (2019)
Useful for intraomics studies and associating them with phenotypes	OMICSONE	Public proteomics data set of ovarian cancer	The program clustered proteins into four clusters, showing association between with a specific protein signature and subtype	R, Python and Jupyter	User friendly, simple to use, and presents data visually	Hu et al., (2018)
Associates frequent aberrations in single omics and multiomics data and establishes underlying mechanisms	OMICS ARULES	DNA methylation and RNA-seq datasets from breast invasive carcinoma (BRCA), esophageal carcinoma (ESCA), and lung adenocarcinoma (LUAD)	The program provides mechanistic links between cause and disease and offers experimentally verifiable hypothesis	R program	This method performs better than other rule ranking functions and identified biologically significant relationships	Chen et al., (2019)

MiRNA interactome in real-time studies can provide a snapshot of the significant pathways being altered during initiation of tumorigenesis, progression to metastasis, making it a strong tool for personalized medicine. It could also be used in single-cell omics to characterize individual cells in that heterogeneous bulk of the tumor, equipping us with the knowledge to target resistant cells in that bulk tumor. With a growing interest in single-cell multiomics and its applications, several reviews have covered the developments (Ma et al., 2020; Chappell et al., 2018; Hu et al., 2018), but lack any specific methodology to cover post-transcriptional regulation at the single-cell level.

Conclusions

We have come a long way from predicting one target for one miRNA and experimentally verifying those using reporter assays to large-scale studies that identify bulk targets and their other regulating partners. There is ongoing research to overcome the difficulties in data integration with upcoming better tools for data integration and interpretation. This evergrowing list of tools indicates the dominating role of miRNAs in physiology, and patholog and the necessity to understand the full picture is growing even more as personalized therapy becomes the norm (Detassis et al., 2017).

This can be accomplished only by multiomics encompassing proteomics. MiRNAs from the huge network of interactome can act as biomarkers, therapy resistance, and prognostic markers, as well as therapeutic targets, and work well with conventional therapy. The reducing costs of omics technologies can go all the way in making the miRNAs a deciding factor for personalized and cancer therapy.

Footnotes

Acknowledgment

We thank for the research fellowships from the Ministry of Human Resource Development, Government of India, to Ms. K J Sindhu, and Indian Council of Medical Research to Dr. Venkatesan Nalini (project number BMS/4566/2019).

Author Disclosure Statement

The authors declare that no competing financial interests exist.

Funding Information

No funding was received for this work.

Abbreviations Used

References

Ab Mutalib

N-S

, Sulaiman

, and Jamal

. (2019). Computational tools for microrna target prediction. In: Computational Epigenetics and Diseases in Translational Epigenetics. Loo Keat Kei, ed. 1st ed., Cambridge, MA: Academic, Press, 79–105.

Advani

, Subbannayya

, Patel

, et al. (2017). Long-term cigarette smoke exposure and changes in mirna expression and proteome in non-small-cell lung cancer. OMICS, 21, 390–403.

Aebersold

, and Mann

. (2003). Mass spectrometry-based proteomics. Nature, 422, 198–207.

Aguet

, Brown

, Castel

, et al. (2017). Genetic effects on gene expression across human tissues. Nature, 550, 204–213.

Akyay

, Gov

, Kenar

, et al. (2020) Mapping the molecular basis and markers of papillary thyroid carcinoma progression and metastasis using global transcriptome and microRNA profiling. OMICS, 24, 148–159.

Ali

, Khan

, Wennerberg

, and Aittokallio

. (2018). Systems biology global proteomics profiling improves drug sensitivity prediction: Results from a multi- omics, pan-cancer modeling approach. Bioinformatics, 34, 1353–1362.

Allendes Osorio

, Nyström-Persson

, Nojima

, Kosugi

, Mizuguchi

, and Natsume-Kitatani

. (2020). Panomicon: A web-based environment for interactive, visual analysis of multi-omics data. Heliyon, 6, e04618.

Aure

, Idbaih

, Dehais

, et al. (2016). Integrated multi-omics analysis of oligodendroglial tumours identifies three subgroups of 1p/19q co-deleted gliomas. Nat Commun, 7, 11263.

Bauer

, and Hummon

. (2012). Effects of the mir-143/-145 microrna cluster on the colon cancer proteome and transcriptome. J Proteome Res, 11, 4744–4754.

10.

Bazzini

, Kee

, and Giraldez

. (2012). Ribosome profiling shows that mir-430 reduces translation before causing mrna decay in zebrafish. Science (80), 336, 233–236.

11.

Berg

KCG

, Eide

, Eilertsen

, et al. (2017). Multi-omics of 34 colorectal cancer cell lines - a resource for biomedical studies. Mol Cancer, 16, 116.

12.

Berges-soria

, Arevalillo

, Nanni

, et al. (2015). Combined label-free quantitative proteomics and microrna expression analysis of breast cancer unravel molecular differences with clinical implications. Cancer Res, 75, 2243–2254.

13.

Bhome

, Vecchio F

Del

, Lee

, et al. (2018). Exosomal micrornas (exomirs): Small molecules with a big role in cancer. Cancer Lett, 420, 228–235.

14.

Bosquet

, Marchion

, Chon

, Lancaster

, and Chanock

. (2014). Analysis of chemotherapeutic response in ovarian cancers using publicly available high-throughput data. Ther Chem Biol, 74, 3902–3913.

15.

Brien

, Hayder

, Zayed

, and Peng

. (2018). Overview of microrna biogenesis, mechanisms of actions, and circulation. Front Endocrinol (Lausanne), 9, 402.

16.

Cadoni

, Manicardi

, and Madder

. (2020). PNA-based microrna detection methodologies. Molecules, 25, 1296.

17.

Castro-Vega

, Letouzé

, Burnichon

, et al. (2015). Multi-omics analysis defines core genomic alterations in pheochromocytomas and paragangliomas. Nat Commun, 6, 6044.

18.

Chakraborty

, Hosen

, Shekhar

, and Ahmed

. (2018). Onco-multi-omics approach: A new frontier in cancer research. Biomed Res Int, 2018, 9836256.

19.

Chappell

, Russell

AJC

, and Voet

. (2018). Single-cell (multi) omics technologies. Annu Rev Genomics Hum Genet, 19, 15–41.

20.

Chaudhry

. (2013). Radiation-induced microrna: Discovery, functional analysis, and cancer radiotherapy. J Cell Biochem, 115, 436–449.

21.

Chen

, Zhang

, Zhao

, and Xu

. (2019). An R package for integration of multi-omics datasets via association rules mining. BMC Bioinformatics, 20, 7–9.

22.

Chen

, Heikkinen

, Wang

, Yang

, Sun

, and Wong

. (2019). Trends in the development of mirna bioinformatics tools. Brief Bioinform, 20, 1836–1852.

23.

Chen

, Yu

, Tsang

, et al. (2011). Systematic proteome analysis identifies transcription factor yy1 as a direct target of mir-34a research articles. J Proteome Res, 10, 479–487.

24.

Cheng

, Zhou

, Xie

, et al. (2010). The impact of mir-34a on protein output in hepatocellular carcinoma hepg2 cells. Proteomics, 10, 1557–1572.

25.

Cheng

, Dong

, Zhang

, Zhao

, and Li

. (2018). Recent advances in microrna detection. Analyst, 143, 1758–1774.

26.

Ciafrè

, and Galardi

. (2013). MicroRNAs and rna-binding proteins a complex network of interactions and reciprocal regulations in cancer. RNA Biol, 10, 934–942.

27.

Coarfa

, Fiskus

, Eedunuri

, et al. (2015). Comprehensive proteomic profiling identifies the androgen receptor axis and other signaling pathways as targets of micrornas suppressed in metastatic prostate cancer. Oncogene, 35, 2345–2356.

28.

Cottrell

, Szczesny

, and Djuranovic

. (2017). Translation efficiency is a determinant of the magnitude of mirna-mediated repression. Sci Rep, 7, 14884.

29.

Cui

, Wang

, Yao

, et al. (2019). Circulating micrornas in cancer: Potential and challenge. Front Genet, 10, 626.

30.

Dave

, Ngo

, Pernestig

, et al. (2019). MicroRNA amplification and detection technologies: Opportunities and challenges for point of care diagnostics. Lab Invest, 99, 452–469.

31.

Dean

, Riahi

, and Wong

. (2015). Spatiotemporal dynamics of microrna during epithelial collective cell migration. Biomaterials, 37, 156–163.

32.

Denli

, Tops

BBJ

, Plasterk

RHA

, Ketting

, and Hannon

. (2004). Processing of primary micrornas by the microprocessor complex. Lett Nat, 432, 231–235.

33.

Detassis

, Grasso

, Vescovo V

Del

, and Denti

. (2017). MicroRNAs make the call in cancer personalized medicine. Front Cell Dev Biol, 5, 86.

34.

Dettmer

, Aronov

, and Hammock

. (2007). MASS spectrometry-based metabolomics. Mass Spectrom Rev, 26, 51–78.

35.

Ebrahimi

, and Hashemy

. (2019). MicroRNA-mediated redox regulation modulates therapy resistance in cancer cells: Clinical perspectives. Cell Oncol, 42, 131–141.

36.

Eichhorn

, Guo

, Mcgeary

, et al. (2010). MRNA destabilization is the dominant effect of mammalian micrornas by the time substantial repression ensues. Mol Cell, 56, 104–115.

37.

Erhart

, Hackl

, Hahne

, et al. (2020). Combined proteomics/mirnomics of dendritic cell immunotherapy-treated glioblastoma patients as a screening for survival-associated factors. Npj Vaccines, 5, 5.

38.

Forero

, Gonzalez-Giraldo

, Castro-Vega

, and Barreto

. (2019). QPCR-based methods for expression analysis of mirnas. Biotechniques, 67, 192–199.

39.

German

, Pillay

, Jeong

, et al. (2008). Global identification of microrna-target rna pairs by parallel analysis of rna ends. Nat Biotechnol, 26, 941–946.

40.

Giambruno

, Mihailovich

, and Bonaldi

. (2018). Mass spectrometry-based proteomics to unveil the non-coding rna world. Front Mol Biosci, 5, 90.

41.

Guo

, Wang

, Yan

, et al. (2020). The role of exosomal microrna in cancer drug resistance. Front Oncol, 10, 472.

42.

Hafner

, Landthaler

, Burger

, et al. (2010). Resource transcriptome-wide identification of rna-binding protein and microrna target sites by par-clip. Cell, 141, 129–141.

43.

Han

, Lee

, Yeom

, Kim

, Jin

, and Kim

. (2004). The drosha – dgcr8 complex in primary microrna processing. Gene Dev, 18, 3016–3027.

44.

Hasin

, Seldin

, and Lusis

. (2017). Multi-omics approaches to disease. Genome Biol, 18, 83.

45.

Helwak

, Kudla

, Dudnakova

, and Tollervey

. (2013). Mapping the human mirna interactome by clash reveals frequent noncanonical binding. Cell, 153, 654–665.

46.

Hosseinahli

, Aghapour

, Baradaran

, and Duijf

PHG

. (2018). Treating cancer with microrna replacement therapy: A literature review. J Cell Physiol, 233, 5574–5588.

47.

Hrovatin

, and Kunej

. (2018). Classification of mirna-related sequence variations. Epigenomics, 10, 463–481.

48.

, An

, Sheu

, Trejo

, Fan

, and Guo

. (2018). Single cell multi-omics technology: Methodology and application. Front Cell Dev Biol, 6, 28.

49.

, Ao

, and Zhang

. (2019). Omics one: Associate omics data with phenotypes in one-click. BioRx iv. 756544.

50.

Huang

, Chaudhary

, and Garmire

. (2017). More is better: Recent progress in multi-omics data integration methods. Front Genet, 8, 84.

51.

Huang

, Renuse

, Pinto

, et al. (2015). Identification of mir-145 targets through an integrated omics analysis. Mol Biosyst, 11, 197–207.

52.

Isenovic

, Petrovic

, and Ergu

. (2017). Levels of microrna heterogeneity in cancer biology. Mol Diagn Ther, 21, 511–523.

53.

Johora

, Hossain

, and Kocerha

. (2019). The potential for microrna therapeutics and clinical research. Front Genet, 10, 478.

54.

Kamal

NNSBNM

, and Shahidan

WNS

. (2020). Non-exosomal and exosomal circulatory micrornas: Which are more valid as biomarkers?. Front Pharmacol, 10, 1500–1522.

55.

Kawamata

, and Tomari

. (2010). Making risc. Trends Biochem Sci, 35, 368–376.

56.

Kazimierczyk

, Kasprowicz

, Kasprzyk

, and Wrzesinski

. (2020). Human long noncoding rna interactome: Detection, characterization and function. Int J Mol Sci, 21, 1027.

57.

Kedde

, Kouwenhove M

Van

, Zwart

, Oude Vrielink

JAF

, Elkon

, and Agami

. (2010). A pumilio-induced rna structure switch in p27-3′2 utr controls mir-221 and mir-222 accessibility. Nat Cell Biol, 12, 1014–1020.

58.

Keene

, Komisarow

, and Friedersdorf

. (2006). RIP-chip: The isolation and identification of mrnas, micrornas and protein components of ribonucleoprotein complexes from cell extracts. Nat Protoc, 1, 302–308.

59.

Kern

, Backes

, Hirsch

, Fehlmann

, Meese

, and Keller

. (2019). What’ s the target: Understanding two decades of in silico microrna-target prediction. Brief Bioinform, 21, 1999–2010.

60.

Kim

, Kim

, Chang

, and You

. (2019). Bias-minimized quantification of microrna reveals widespread alternative processing and 3 end modification. Nucleic Acids Res, 47, 2630–2640.

61.

Kobayashi

, Ono

, Satow

, et al. (2014). MicroRNA expression and functional profiles of osteosarcoma. Clin Transl Res, 86, 94–103.

62.

Kong

, Yao

, Bing

, et al. (2020). Dynamical network analysis reveals key micrornas in progressive stages of lung cancer. PLoS Comput Biol, 16, 1–23.

63.

Kong

, Qiao

, Ren

, et al. (2020). Integrative analysis of membrane proteome and microrna reveals novel lung cancer metastasis biomarkers. Front Genet, 11, 1023.

64.

Kulski

. (2015). Next-generation sequencing—an overview of the history, tools, and “omic” applications. In: Next Generation Sequencing Advances, Applications and Challenges. Kulski, JK, ed. London:, IntechOpen, 1–59.

65.

Kwon

, Kim

, Lee

, et al. (2015). Integrative analysis of multi-omics data for identifying multi-markers for diagnosing pancreatic cancer. BMC Genomics, 16, 88.

66.

Laakso

, Kuusisto

, Stančáková

, et al. (2017). The metabolic syndrome in men study: A resource for studies of metabolic & cardiovascular diseases. J Lipid Res, 58, 481–493.

67.

Lee

, Feinbaum

, and Ambros

. (1993). The C. elegans heterochronic gene lin-4 encodes small rnas with antisense complementarity to lin-14. Cell, 75, 843–854.

68.

Lee

, Ahn

, Han

, et al. (2003). The nuclear rnase iii drosha initiates microrna processing. Nature, 425, 415–419.

69.

, Xiong

, Ge

, and Bi

. (2012). Quantitative proteomic strategies for the identification of microrna targets. Expert Rev Proteomics, 9, 549–559.

70.

, and Zhang

. (2019). Current experimental strategies for intracellular target identification of microrna. ExRNA, 1, 6.

71.

, Min

, and Zhang

. (2009). A probabilistic framework to improve microrna target prediction by incorporating proteomics data. J Bioinform Comput Biol, 7, 955–972.

72.

Licatalosi

, Mele

, Fak

, et al. (2008). HITS-clip yields genome-wide insights into brain alternative rna processing. Nature, 456, 464–470.

73.

Liu

, Halvey

, Shyr

, Slebos

RJC

, Liebler

, and Zhang

. (2013). Integrative omics analysis reveals the importance and scope of translational repression in microrna-mediated regulation. Mol Cell Proteomics, 12, 1900–1911.

74.

Loher

, Londin

, and Rigoutsos

. (2014). IsomiR expression profiles in human lymphoblastoid cell lines exhibit population and gender dependencies. Oncotarget, 5, 8790–8802.

75.

López-urrutia

, Montes

LPB

, Ladrón

, Cervantes

, and Pérez-plasencia

. (2019). Crosstalk between long non-coding rnas, micro-rnas and mrnas: Deciphering molecular mechanisms of master regulators in cancer. Front Oncol, 9, 669.

76.

, and Zhan

. (2018). The crucial role of multiomic approach in cancer research and clinically relevant outcomes. EPMA J, 9, 77–102.

77.

, and Leslie

. (2016). Learning to predict mirna-mrna interactions from ago clip sequencing and clash data. PLoS Comput Biol, 12, e1005026.

78.

Luo

, Dai

, Yin

, et al. (2015). Multiple reaction monitoring-ion pair finder: A systematic approach to transform nontargeted mode to pseudotargeted mode for metabolomics study based on liquid chromatography − mass spectrometry. Anal Chem, 87, 5050–5055.

79.

, Mcdermaid

, Xu

, and Chang

. (2020). Integrative methods and practical challenges for single-cell multi-omics. Trends Biotechnol, 38, 1007–1022.

80.

, Zhu

, Huang

, Tegeler

, Gao

, and Zhang

. (2015). Quantitative proteomic approach for microrna target prediction based on 18 o/16 o labeling. Cancer Inform, 14, 163–173.

81.

Madhavan

, Gusev

, Natarajan

, et al. (2013). Genome-wide multi-omics profiling of colorectal cancer identifies immune determinants strongly associated with relapse. Front Genet, 4, 236.

82.

Marta

, Garicochea B

C L

A, R M

, and P K

. (2015). MicroRNAs, cancer and ionizing radiation: Where are we?. Rev Assoc Med Bras, 61, 275–281.

83.

Matsiko

. (2018). Cancer immunotherapy making headway. Nat Mater, 17, 472.

84.

Misra

, Misra

, Langefeld

, Olivier

, and Cox

. (2019). Integrated omics: Tools, advances and future approaches. J Mol Endocrinol, 62, R21–R45.

85.

Mitra

, Adams

, Jiang

, Greenawalt

, and Eischen

. (2020). Pan-cancer analysis reveals cooperativity of both strands of microrna that regulate tumorigenesis and patient survival. Nat Commun, 11, 968.

86.

Moertl

, Mutschelknaus

, Heider

, and Atkinson

. (2016). MicroRNAs as novel elements in personalized radiotherapy. Transl Cancer Res, 5, 1262–1269.

87.

Mognato

, and Celotti

. (2015). MicroRNAs used in combination with anti-cancer treatments can enhance therapy efficacy. Med Chem (Los. Angeles), 15, 1052–1062.

88.

Mollaei

, Safaralizadeh

, and Rostami

. (2019). MicroRNA replacement therapy in cancer. J Cell Physiol, 234, 12369–12384.

89.

Moore

, Scheel

TKH

, Luna

, et al. (2015). MiRNA–target chimeras reveal mirna 30-end pairing as a major determinant of argonaute target specificity. Nat Commun, 6, 8864.

90.

Nagarajan

, Tentori

, Zhang

, Slack

, and Doyle

. (2020). Spatially resolved and multiplexed microrna quantification from tissue using nanoliter well arrays. Microsystems Nanoeng, 6, 51.

91.

Nassa

, Tarallo

, Giurato

, et al. (2014). Post-transcriptional regulation of human breast cancer cell proteome by unliganded estrogen receptor via micrornas. Mol Cell Proteomics, 13, 1076–1090.

92.

Nguyen

, and Chang

. (2018). Development of novel therapeutic agents by inhibition of oncogenic micrornas. Int J Mol Sci, 19, 65.

93.

Nguyen

, Shrestha

, Draghici

, et al. (2019). PINSPlus: A tool for tumor subtype discovery in integrated genomic data. Bioinformatics, 35, 2843–2846.

94.

Nouri-Nigjeh

, Chen

, and Pan

. (2015). Targeted proteomics in translational and clinical studies. In: Recent Advances in Proteomics Research. Magdeldin S, ed. London:, IntechOpen, 135–147.

95.

Nussbacher

, and Yeo

. (2018). Systematic discovery of rna binding proteins that regulate microrna levels article systematic discovery of rna binding proteins that regulate microrna levels. Mol Cell, 69, 1005–1016.

96.

Ors-kumoglu

, Gluce-Iz

, and Biray-Avci

. (2019). Therapeutic micrornas in human cancer. Cytotechnology, 71, 411–425.

97.

Pasquini

, and Kunej

. (2019). A map of the microrna regulatory networks identified by experimentally validated microrna-target interactions in five domestic animals: Cattle, pig, sheep, dog, and chicken. OMICS, 23, 448–456.

98.

Peng

, and Croce

. (2016). The role of micrornas in human cancer. Signal Transduct Target Ther, 1, 15004.

99.

Piletic

, and Kunej

. (2018). MicroRNA-target interactions reloaded: Identification of potentially functional sequence variants within validated microrna-target interactions. OMICS, 22, 700–708.

100.

Pouryahya

, Oh

, Javanmard

, et al. (2020). aWCluster: A novel integrative network-based clustering of multi-omics for subtype analysis of cancer data. IEEE/ACM Trans Comput Biol Bioinformatics, 1, PP1-1, 5555.

101.

Qin

, Li

, Wang

, et al. (2013). ProteoMirExpress: Inferring microrna and protein-centered regulatory networks from high-throughput proteomic and mrna expression data. Mol Cell Proteomics, 12, 3379–3387.

102.

Rappoport

, Shamir

, and Schwartz

. (2019). NEMO: Cancer subtyping by integration of partial multi-omics data. Bioinformatics, 35, 3348–3356.

103.

Reczko

, Maragkakis

, Alexiou

, Papadopoulos

, and Hatzigeorgiou

. (2012). Accurate microrna target prediction using detailed binding site accessibility and machine learning on proteomics data. Front Genet, 2, 103.

104.

Sathyanarayanan

, Gupta

, Thompson

, Nyholt

, Bauer

, and Nagaraj

. (2019). A comparative study of multi-omics integration tools for cancer driver gene identification and tumour subtyping. Brief Bioinform, 21, 1920–1936.

105.

Sen

, Ghosal

, Das

, Balti

, and Chakrabarti

. (2014). Competing endogenous rna: The key to posttranscriptional regulation. Sci World J, 2014, 896206.

106.

Seviour

, Sehgal

, Lu

, et al. (2016). Functional proteomics identifies mirnas to target a p27/myc/phospho-rb signature in breast and ovarian cancer. Oncogene, 35, 691–701.

107.

, Shen

, Zheng

, and Fan

. (2019). The role and mechanisms of action of micrornas in cancer drug resistance. Clin Epigenetics, 11, 25.

108.

Subramanian

, Verma

, Kumar

, Jere

, and Anamika

. (2020). Multi-omics data integration, interpretation, and its application. Bioinform Biol Insights, 14, 7–9.

109.

Sun

, Shi

, Yang

, et al. (2018). Effect of exosomal mirna on cancer biology and clinical applications. Mol Cancer, 17, 147.

110.

Svoronos

, Engelman

, and Slack

. (2016). OncomiR or tumor suppressor? the duplicity of micrornas in cancer. Cancer Res, 76, 3666–3670.

111.

Tan

, Huang

, Zhang

, Qian

, Sun

, and Zhou

. (2019). EBioMedicine pan-cancer analysis on microrna-associated gene activation. EBioMedicine, 43, 82–97.

112.

Tang

, Lee

, Sharpe

, et al. (2012). Oxidative stress-responsive microrna-320 regulates glycolysis in diverse biological systems. FASEB J, 26, 4710–4721.

113.

Telonis

, Loher

, Jing

, Londin

, and Rigoutsos

. (2015). Beyond the one-locus-one-mirna paradigm: Microrna isoforms enable deeper insights into breast cancer heterogeneity. Nucleic Acids Res, 43, 9158–9175.

114.

Telonis

, Magee

, Loher

, Chervoneva

, Londin

, and Rigoutsos

. (2017). Knowledge about the presence or absence of mirna isoforms (isomirs) can successfully discriminate amongst 32 tcga cancer types. Nucleic Acids Res, 45, 2973–2985.

115.

Torres

, Garcia-palmero

, Bartolomé

, Fernandez-aceñero

, and Molina

. (2018). Combined mirna profiling and proteomics demonstrates that different mirnas target a common set of proteins to promote colorectal cancer metastasis: Mirnas in colorectal cancer meta … combined mirna profiling and proteomics demonstrates that different m. J Pathol, 242, 39–51.

116.

Turunen

, Roberts

, Laitinen

, et al. (2019). Changes in nuclear and cytoplasmic microrna distribution in response to hypoxic stress. Sci Rep, 9, 10332.

117.

Uhlmann

, Zhang

, and Schwa

. (2010). MiR-200bc/429 cluster targets plc c 1 and differentially regulates proliferation and egf-driven invasion than mir-200a/141 in breast cancer. Oncogene, 29, 4297–4306.

118.

Uhlmann

, Mannsperger

, Zhang

, et al. (2012). Global microrna level regulation of egfr-driven cell-cycle protein network in breast cancer. Mol Syst Biol, 8, 570.

119.

Ulfenborg

. (2019). Vertical and horizontal of multi-omics with Miodin. BMC Bioinformatics, 20, 1–10.

120.

Wang

, and Jing Huan

. (2012). RNA helicase ddx5 regulates microrna expression and contributes to cytoskeletal reorganization in basal breast cancer cells. Mol Cell Proteomics, 11, M111.011932.

121.

Wang

, Peng

, Wang

, Qin

, and Xue

. (2018). Circulating micrornas as potential cancer biomarkers: The advantage and disadvantage. Clin Epigenetics, 10, 59.

122.

Wei

, Zhang

, Weng

, Chen

, and Cai

. (2020). Survey and comparative assessments of computational multi-omics integrative methods with multiple regulatory networks identifying distinct tumor compositions across pan-cancer data sets. Brief Bioinform, 22, bbaa102.

123.

Weinstein

, Collisson

, Mills

, et al. (2013). The cancer genome atlas pan-cancer analysis project. Nat Genet, 45, 1113–1120.

124.

Wheeler

and Roberts

. (2017). Comprehensive and integrative genomic characterization of hepatocellular carcinoma resource comprehensive and integrative genomic characterization of hepatocellular carcinoma. Cell, 169, 1327–1341.

125.

Wiench

, Chen

, Paulsen

, et al. (2013). Integration of different “‘-omics’”. technologies identifies inhibition of the igf1r-akt-mtor signaling cascade involved in the cytotoxic effect of shikonin against leukemia cells. Evidence-Based Complement. Altern 11, 818709.

126.

Willmann

, Berkowitz

, and Gregory

. (2014). Improved genome-wide mapping of uncapped and cleaved transcripts in eukaryotes-gmuct 2.0. Methods, 67, 64–73.

127.

Xie

, Yuan

, Yang

, Sun

, Zhou

, and Fang

. (2018). MOBCdb: A comprehensive database integrating multi-omics data on breast cancer for precision medicine. Breast Cancer Res Treat, 169, 625–632.

128.

, Yang

, and Chen

. (2015). Quantification of microrna by dna−peptide probe and liquid chromatography -tandem mass spectrometry based quasi targeted proteomics. Anal Chem, 88, 754–763.

129.

Yang

, Chaerkady

, Kandasamy

, Huang

, Mendell

, and Pandey

. (2010). Identifying targets of mir-143 using a silac-based proteomic approach. Mol Biosyst, 6, 1873–1882.

130.

Yang

, Liu

, Lin

, Zhang

, and Wang

. (2019). Multiomics analysis on dna methylation and the expression of both messenger rna and microrna in lung adenocarcinoma. J Cell Physiol, 234, 7579–7586.

131.

Yoo

, Kim

, Woo

, and Myung

. (2018). Clinical multi-omics strategies for the effective cancer management. J Proteomics, 188, 97–106.

132.

Yoon

, Srikantan

, and Gorospe

. (2012). MS2-trap (ms2-tagged rna affinity purification): Tagging rna to identify associated mirnas. Methods, 58, 81–87.

133.

Zaheer

, Faheem

, Qadri

, et al. (2019). Expression profile of microrna: An emerging hallmark of cancer. Curr Pharm Des, 25, 642–653.

134.

Zealy

, Wrenn

, Davila

, Min

, and Yoon

. (2017). MicroRNA binding proteins:Speci ficity and function. WIREs RNA, 8, e1414.

135.

Zhang

, Kolb

, Jaskiewicz

, Westhof

, and Filipowicz

. (2004). Single processing center models for human dicer and bacterial rnase iii. Cell, 118, 57–68.

136.

Zhang

, Li

, Liu

, et al. (2016a). Role of plant microrna in cross-species regulatory networks of humans. BMC Syst Biol, 10, 60.

137.

Zhang

, Huang

, Gu

, et al. (2016b). Molecular subtyping of serous ovarian cancer based on multi-omics data. Sci Rep, 6, 26001.