Downregulation of Dystrophin Expression Occurs across Diverse Tumors, Correlates with the Age of Onset, Staging and Reduced Survival of Patients.

Altered dystrophin expression was found in some tumors and recent studies identified a developmental onset of Duchenne muscular dystrophy (DMD). Given that embryogenesis and carcinogenesis share many mechanisms, we analyzed a broad spectrum of tumors to establish whether dystrophin alteration evokes related outcomes. Transcriptomic, proteomic, and mutation datasets from fifty tumor tissues and matching controls (10,894 samples) and 140 corresponding tumor cell lines were analyzed. Interestingly, dystrophin transcripts and protein expression were found widespread across healthy tissues and at housekeeping gene levels. In 80% of tumors, DMD expression was reduced due to transcriptional downregulation and not somatic mutations. The full-length transcript encoding Dp427 was decreased in 68% of tumors, while Dp71 variants showed variability of expression. Notably, low expression of dystrophins was associated with a more advanced stage, older age of onset, and reduced survival across different tumors. Hierarchical clustering analysis of DMD transcripts distinguished malignant from control tissues. Transcriptomes of primary tumors and tumor cell lines with low DMD expression showed enrichment of specific pathways in the differentially expressed genes. Pathways consistently identified: ECM-receptor interaction, calcium signaling, and PI3K-Akt are also altered in DMD muscle. Therefore, the importance of this largest known gene extends beyond its roles identified in DMD, and certainly into oncology.

Predicting genes associated with RNA methylation pathways using machine learning.

RNA methylation plays an important role in functional regulation of RNAs, and has thus attracted an increasing interest in biology and drug discovery. Here, we collected and collated transcriptomic, proteomic, structural and physical interaction data from the Harmonizome database, and applied supervised machine learning to predict novel genes associated with RNA methylation pathways in human. We selected five types of classifiers, which we trained and evaluated using cross-validation on multiple training sets. The best models reached 88% accuracy based on cross-validation, and an average 91% accuracy on the test set. Using protein-protein interaction data, we propose six molecular sub-networks linking model predictions to previously known RNA methylation genes, with roles in mRNA methylation, tRNA processing, rRNA processing, but also protein and chromatin modifications. Our study exemplifies how access to large omics datasets joined by machine learning methods can be used to predict gene function.

dialogi: Utilising NLP With Chemical and Disease Similarities to Drive the Identification of Drug-Induced Liver Injury Literature.

Drug-Induced Liver Injury (DILI), despite its low occurrence rate, can cause severe side effects or even lead to death. Thus, it is one of the leading causes for terminating the development of new, and restricting the use of already-circulating, drugs. Moreover, its multifactorial nature, combined with a clinical presentation that often mimics other liver diseases, complicate the identification of DILI-related (or "positive") literature, which remains the main medium for sourcing results from the clinical practice and experimental studies. This work-contributing to the "Literature AI for DILI Challenge" of the Critical Assessment of Massive Data Analysis (CAMDA) 2021- presents an automated pipeline for distinguishing between DILI-positive and negative publications. We used Natural Language Processing (NLP) to filter out the uninformative parts of a text, and identify and extract mentions of chemicals and diseases. We combined that information with small-molecule and disease embeddings, which are capable of capturing chemical and disease similarities, to improve classification performance. The former were directly sourced from the Chemical Checker (CC). For the latter, we collected data that encode different aspects of disease similarity from the National Library of Medicine's (NLM) Medical Subject Headings (MeSH) thesaurus and the Comparative Toxicogenomics Database (CTD). Following a similar procedure as the one used in the CC, vector representations for diseases were learnt and evaluated. Two Neural Network (NN) classifiers were developed: a baseline model that accepts texts as input and an augmented, extended, model that also utilises chemical and disease embeddings. We trained, validated, and tested the classifiers through a Nested Cross-Validation (NCV) scheme with 10 outer and 5 inner folds. During this, the baseline and extended models performed virtually identically, with F1-scores of 95.04 ± 0.61% and 94.80 ± 0.41%, respectively. Upon validation on an external, withheld, dataset that is meant to assess classifier generalisability, the extended model achieved an F1-score of 91.14 ± 1.62%, outperforming its baseline counterpart which received a lower score of 88.30 ± 2.44%. We make further comparisons between the classifiers and discuss future improvements and directions, including utilising chemical and disease embeddings for visualisation and exploratory analysis of the DILI-positive literature.

DILI C : An AI-Based Classifier to Search for Drug-Induced Liver Injury Literature.

Drug-induced liver injury (DILI) is a class of adverse drug reactions (ADR) that causes problems in both clinical and research settings. It is the most frequent cause of acute liver failure in the majority of Western countries and is a major cause of attrition of novel drug candidates. Manual trawling of the literature is the main route of deriving information on DILI from research studies. This makes it an inefficient process prone to human error. Therefore, an automatized AI model capable of retrieving DILI-related articles from the huge ocean of literature could be invaluable for the drug discovery community. In this study, we built an artificial intelligence (AI) model combining the power of natural language processing (NLP) and machine learning (ML) to address this problem. This model uses NLP to filter out meaningless text (e.g., stop words) and uses customized functions to extract relevant keywords such as singleton, pair, and triplet. These keywords are processed by an apriori pattern mining algorithm to extract relevant patterns which are used to estimate initial weightings for a ML classifier. Along with pattern importance and frequency, an FDA-approved drug list mentioning DILI adds extra confidence in classification. The combined power of these methods builds a DILI classifier (DILI C ), with 94.91% cross-validation and 94.14% external validation accuracy. To make DILI C as accessible as possible, including to researchers without coding experience, an R Shiny app capable of classifying single or multiple entries for DILI is developed to enhance ease of user experience and made available at https://researchmind.co.uk/diliclassifier/. Additionally, a GitHub link (https://github.com/sanjaysinghrathi/DILI-Classifier) for app source code and ISMB extended video talk (https://www.youtube.com/watch?v=j305yIVi_f8) are available as supplementary materials.

Computational Discovery of Cancer Immunotherapy Targets by Intercellular CRISPR Screens.

Cancer immunotherapy targets the interplay between immune and cancer cells. In particular, interactions between cytotoxic T lymphocytes (CTLs) and cancer cells, such as PD-1 (PDCD1) binding PD-L1 (CD274), are crucial for cancer cell clearance. However, immune checkpoint inhibitors targeting these interactions are effective only in a subset of patients, requiring the identification of novel immunotherapy targets. Genome-wide clustered regularly interspaced short palindromic repeats (CRISPR) screening in either cancer or immune cells has been employed to discover regulators of immune cell function. However, CRISPR screens in a single cell type complicate the identification of essential intercellular interactions. Further, pooled screening is associated with high noise levels. Herein, we propose intercellular CRISPR screens, a computational approach for the analysis of genome-wide CRISPR screens in every interacting cell type for the discovery of intercellular interactions as immunotherapeutic targets. We used two publicly available genome-wide CRISPR screening datasets obtained while triple-negative breast cancer (TNBC) cells and CTLs were interacting. We analyzed 4825 interactions between 1391 ligands and receptors on TNBC cells and CTLs to evaluate their effects on CTL function. Intercellular CRISPR screens discovered targets of approved drugs, a few of which were not identifiable in single datasets. To evaluate the method's performance, we used data for cytokines and costimulatory molecules as they constitute the majority of immunotherapeutic targets. Combining both CRISPR datasets improved the recall of discovering these genes relative to using single CRISPR datasets over two-fold. Our results indicate that intercellular CRISPR screens can suggest novel immunotherapy targets that are not obtained through individual CRISPR screens. The pipeline can be extended to other cancer and immune cell types to discover important intercellular interactions as potential immunotherapeutic targets.

Deriving time-concordant event cascades from gene expression data: A case study for Drug-Induced Liver Injury (DILI).

Adverse event pathogenesis is often a complex process which compromises multiple events ranging from the molecular to the phenotypic level. In toxicology, Adverse Outcome Pathways (AOPs) aim to formalize this as temporal sequences of events, in which event relationships should be supported by causal evidence according to the tailored Bradford-Hill criteria. One of the criteria is whether events are consistently observed in a certain temporal order and, in this work, we study this time concordance using the concept of "first activation" as data-driven means to generate hypotheses on potentially causal mechanisms. As a case study, we analysed liver data from repeat-dose studies in rats from the TG-GATEs database which comprises measurements across eight timepoints, ranging from 3 hours to 4 weeks post-treatment. We identified time-concordant gene expression-derived events preceding adverse histopathology, which serves as surrogate readout for Drug-Induced Liver Injury (DILI). We find known mechanisms in DILI to be time-concordant, and show further that significance, frequency and log fold change (logFC) of differential expression are metrics which can additionally prioritize events although not necessary to be mechanistically relevant. Moreover, we used the temporal order of transcription factor (TF) expression and regulon activity to identify transcriptionally regulated TFs and subsequently combined this with prior knowledge on functional interactions to derive detailed gene-regulatory mechanisms, such as reduced Hnf4a activity leading to decreased expression and activity of Cebpa. At the same time, also potentially novel events are identified such as Sox13 which is highly significantly time-concordant and shows sustained activation over time. Overall, we demonstrate how time-resolved transcriptomics can derive and support mechanistic hypotheses by quantifying time concordance and how this can be combined with prior causal knowledge, with the aim of both understanding mechanisms of toxicity, as well as potential applications to the AOP framework. We make our results available in the form of a Shiny app (https://anikaliu.shinyapps.io/dili_cascades), which allows users to query events of interest in more detail.

Identification of potential pan-coronavirus therapies using a computational drug repurposing platform.

In the past 20 years, there have been several infectious disease outbreaks in humans for which the causative agent has been a zoonotic coronavirus. Novel infectious disease outbreaks, as illustrated by the current coronavirus disease 2019 (COVID-19) pandemic, demand a rapid response in terms of identifying effective treatments for seriously ill patients. The repurposing of approved drugs from other therapeutic areas is one of the most practical routes through which to approach this. Here, we present a systematic network-based drug repurposing methodology, which interrogates virus-human, human protein-protein and drug-protein interactome data. We identified 196 approved drugs that are appropriate for repurposing against COVID-19 and 102 approved drugs against a related coronavirus, severe acute respiratory syndrome (SARS-CoV). We constructed a protein-protein interaction (PPI) network based on disease signatures from COVID-19 and SARS multi-omics datasets. Analysis of this PPI network uncovered key pathways. Of the 196 drugs predicted to target COVID-19 related pathways, 44 (hypergeometric p-value: 1.98e-04) are already in COVID-19 clinical trials, demonstrating the validity of our approach. Using an artificial neural network, we provide information on the mechanism of action and therapeutic value for each of the identified drugs, to facilitate their rapid repurposing into clinical trials.

Identification of SARS-CoV-2-induced pathways reveals drug repurposing strategies.

The global outbreak of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) necessitates the rapid development of new therapies against coronavirus disease 2019 (COVID-19) infection. Here, we present the identification of 200 approved drugs, appropriate for repurposing against COVID-19. We constructed a SARS-CoV-2-induced protein network, based on disease signatures defined by COVID-19 multiomics datasets, and cross-examined these pathways against approved drugs. This analysis identified 200 drugs predicted to target SARS-CoV-2-induced pathways, 40 of which are already in COVID-19 clinical trials, testifying to the validity of the approach. Using artificial neural network analysis, we classified these 200 drugs into nine distinct pathways, within two overarching mechanisms of action (MoAs): viral replication (126) and immune response (74). Two drugs (proguanil and sulfasalazine) implicated in viral replication were shown to inhibit replication in cell assays. This unbiased and validated analysis opens new avenues for the rapid repurposing of approved drugs into clinical trials.

Methylation of histone H3 at lysine 37 by Set1 and Set2 prevents spurious DNA replication.

DNA replication initiates at genomic locations known as origins of replication, which, in S. cerevisiae, share a common DNA consensus motif. Despite being virtually nucleosome-free, origins of replication are greatly influenced by the surrounding chromatin state. Here, we show that histone H3 lysine 37 mono-methylation (H3K37me1) is catalyzed by Set1p and Set2p and that it regulates replication origin licensing. H3K37me1 is uniformly distributed throughout most of the genome, but it is scarce at replication origins, where it increases according to the timing of their firing. We find that H3K37me1 hinders Mcm2 interaction with chromatin, maintaining low levels of MCM outside of conventional replication origins. Lack of H3K37me1 results in defective DNA replication from canonical origins while promoting replication events at inefficient and non-canonical sites. Collectively, our results indicate that H3K37me1 ensures correct execution of the DNA replication program by protecting the genome from inappropriate origin licensing and spurious DNA replication.

Current and prospective computational approaches and challenges for developing COVID-19 vaccines.

SARS-CoV-2, which causes COVID-19, was first identified in humans in late 2019 and is a coronavirus which is zoonotic in origin. As it spread around the world there has been an unprecedented effort in developing effective vaccines. Computational methods can be used to speed up the long and costly process of vaccine development. Antigen selection, epitope prediction, and toxicity and allergenicity prediction are areas in which computational tools have already been applied as part of reverse vaccinology for SARS-CoV-2 vaccine development. However, there is potential for computational methods to assist further. We review approaches which have been used and highlight additional bioinformatic approaches and PK modelling as in silico methods which may be useful for SARS-CoV-2 vaccine design but remain currently unexplored. As more novel viruses with pandemic potential are expected to arise in future, these techniques are not limited to application to SARS-CoV-2 but also useful to rapidly respond to novel emerging viruses.

Inflammatory Signals Induce AT2 Cell-Derived Damage-Associated Transient Progenitors that Mediate Alveolar Regeneration.

Tissue regeneration is a multi-step process mediated by diverse cellular hierarchies and states that are also implicated in tissue dysfunction and pathogenesis. Here we leveraged single-cell RNA sequencing in combination with in vivo lineage tracing and organoid models to finely map the trajectories of alveolar-lineage cells during injury repair and lung regeneration. We identified a distinct AT2-lineage population, damage-associated transient progenitors (DATPs), that arises during alveolar regeneration. We found that interstitial macrophage-derived IL-1ß primes a subset of AT2 cells expressing Il1r1 for conversion into DATPs via a HIF1α-mediated glycolysis pathway, which is required for mature AT1 cell differentiation. Importantly, chronic inflammation mediated by IL-1ß prevents AT1 differentiation, leading to aberrant accumulation of DATPs and impaired alveolar regeneration. Together, this stepwise mapping to cell fate transitions shows how an inflammatory niche controls alveolar regeneration by controlling stem cell fate and behavior.

A novel long noncoding RNA Linc-ASEN represses cellular senescence through multileveled reduction of p21 expression.

Long noncoding RNAs (lncRNAs) regulating diverse cellular processes implicate in many diseases. However, the function of lncRNAs in cellular senescence remains largely unknown. Here we identify a novel long intergenic noncoding RNA Linc-ASEN expresses in prematurely senescent cells. We find that Linc-ASEN associates with UPF1 by RNA pulldown mass spectrometry analysis, and represses cellular senescence by reducing p21 production transcriptionally and posttranscriptionally. Mechanistically, the Linc-ASEN-UPF1 complex suppressed p21 transcription by recruiting Polycomb Repressive Complex 1 (PRC1) and PRC2 to the p21 locus, and thereby preventing binding of the transcriptional activator p53 on the p21 promoter through histone modification. In addition, the Linc-ASEN-UPF1 complex repressed p21 expression posttranscriptionally by enhancing p21 mRNA decay in association with DCP1A. Accordingly, Linc-ASEN levels were found to correlate inversely with p21 mRNA levels in tumors from patient-derived mouse xenograft, in various human cancer tissues, and in aged mice tissues. Our results reveal that Linc-ASEN prevents cellular senescence by reducing the transcription and stability of p21 mRNA in concert with UPF1, and suggest that Linc-ASEN might be a potential therapeutic target in processes influenced by senescence, including cancer.

Differential Expression of Soluble Receptor for Advanced Glycation End-products in Mice Susceptible or Resistant to Chronic Colitis.

BACKGROUND: Identifying the factors that contribute to chronicity in inflamed colitic tissue is not trivial. However, in mouse models of colitis, we can investigate at preclinical timepoints. We sought to validate murine Trichuris muris infection as a model for identification of factors that promote development of chronic colitis. METHODS: We compared preclinical changes in mice with a resolving immune response to T. muris (resistant) vs mice that fail to expel the worms and develop chronic colitis (susceptible). Findings were then validated in healthy controls and patients with suspected or confirmed IBD. RESULTS: The receptor for advanced glycation end products (RAGE) was highly dysregulated between resistant and susceptible mice before the onset of any pathological signs. Increased soluble RAGE (sRAGE) in the serum and feces of resistant mice correlated with reduced colitis scores. Mouse model findings were validated in a preliminary clinical study: fecal sRAGE was differentially expressed in patients with active IBD compared with IBD in remission, patients with IBD excluded, or healthy controls. CONCLUSIONS: Preclinical changes in mouse models can identify early pathways in the development of chronic inflammation that human studies cannot. We identified the decoy receptor sRAGE as a potential mechanism for protection against chronic inflammation in colitis in mice and humans. We propose that the RAGE pathway is clinically relevant in the onset of chronic colitis and that further study of sRAGE in IBD may provide a novel diagnostic and therapeutic target.

Long non-coding RNA ChRO1 facilitates ATRX/DAXX-dependent H3.3 deposition for transcription-associated heterochromatin reorganization.

Constitutive heterochromatin undergoes a dynamic clustering and spatial reorganization during myogenic differentiation. However the detailed mechanisms and its role in cell differentiation remain largely elusive. Here, we report the identification of a muscle-specific long non-coding RNA, ChRO1, involved in constitutive heterochromatin reorganization. ChRO1 is induced during terminal differentiation of myoblasts, and is specifically localized to the chromocenters in myotubes. ChRO1 is required for efficient cell differentiation, with global impacts on gene expression. It influences DNA methylation and chromatin compaction at peri/centromeric regions. Inhibition of ChRO1 leads to defects in the spatial fusion of chromocenters, and mislocalization of H4K20 trimethylation, Suv420H2, HP1, MeCP2 and cohesin. In particular, ChRO1 specifically associates with ATRX/DAXX/H3.3 complex at chromocenters to promote H3.3 incorporation and transcriptional induction of satellite repeats, which is essential for chromocenter clustering. Thus, our results unveil a mechanism involving a lncRNA that plays a role in large-scale heterochromatin reorganization and cell differentiation.

DDX3X RNA helicase affects breast cancer cell cycle progression by regulating expression of KLF4.

DDX3X is a multifunctional RNA helicase with documented roles in different cancer types. Here, we demonstrate that DDX3X plays an oncogenic role in breast cancer cells by modulating the cell cycle. Depletion of DDX3X in MCF7 cells slows cell proliferation by inducing a G1 phase arrest. Notably, DDX3X inhibits expression of Kruppel-like factor 4 (KLF4), a transcription factor and cell cycle repressor. Moreover, DDX3X directly interacts with KLF4 mRNA and regulates its splicing. We show that DDX3X-mediated repression of KLF4 promotes expression of S-phase inducing genes in MCF7 breast cancer cells. These findings provide evidence for a novel function of DDX3X in regulating expression and downstream functions of KLF4, a master negative regulator of the cell cycle.

Genomic positional conservation identifies topological anchor point RNAs linked to developmental loci.

BACKGROUND: The mammalian genome is transcribed into large numbers of long noncoding RNAs (lncRNAs), but the definition of functional lncRNA groups has proven difficult, partly due to their low sequence conservation and lack of identified shared properties. Here we consider promoter conservation and positional conservation as indicators of functional commonality. RESULTS: We identify 665 conserved lncRNA promoters in mouse and human that are preserved in genomic position relative to orthologous coding genes. These positionally conserved lncRNA genes are primarily associated with developmental transcription factor loci with which they are coexpressed in a tissue-specific manner. Over half of positionally conserved RNAs in this set are linked to chromatin organization structures, overlapping binding sites for the CTCF chromatin organiser and located at chromatin loop anchor points and borders of topologically associating domains (TADs). We define these RNAs as topological anchor point RNAs (tapRNAs). Characterization of these noncoding RNAs and their associated coding genes shows that they are functionally connected: they regulate each other's expression and influence the metastatic phenotype of cancer cells in vitro in a similar fashion. Furthermore, we find that tapRNAs contain conserved sequence domains that are enriched in motifs for zinc finger domain-containing RNA-binding proteins and transcription factors, whose binding sites are found mutated in cancers. CONCLUSIONS: This work leverages positional conservation to identify lncRNAs with potential importance in genome organization, development and disease. The evidence that many developmental transcription factors are physically and functionally connected to lncRNAs represents an exciting stepping-stone to further our understanding of genome regulation.

Promoter-bound METTL3 maintains myeloid leukaemia by m6A-dependent translation control.

N6-methyladenosine (m6A) is an abundant internal RNA modification in both coding and non-coding RNAs that is catalysed by the METTL3-METTL14 methyltransferase complex. However, the specific role of these enzymes in cancer is still largely unknown. Here we define a pathway that is specific for METTL3 and is implicated in the maintenance of a leukaemic state. We identify METTL3 as an essential gene for growth of acute myeloid leukaemia cells in two distinct genetic screens. Downregulation of METTL3 results in cell cycle arrest, differentiation of leukaemic cells and failure to establish leukaemia in immunodeficient mice. We show that METTL3, independently of METTL14, associates with chromatin and localizes to the transcriptional start sites of active genes. The vast majority of these genes have the CAATT-box binding protein CEBPZ present at the transcriptional start site, and this is required for recruitment of METTL3 to chromatin. Promoter-bound METTL3 induces m6A modification within the coding region of the associated mRNA transcript, and enhances its translation by relieving ribosome stalling. We show that genes regulated by METTL3 in this way are necessary for acute myeloid leukaemia. Together, these data define METTL3 as a regulator of a chromatin-based pathway that is necessary for maintenance of the leukaemic state and identify this enzyme as a potential therapeutic target for acute myeloid leukaemia.

TIGERi: modeling and visualizing the responses to perturbation of a transcription factor network.

BACKGROUND: Transcription factor (TF) networks play a key role in controlling the transfer of genetic information from gene to mRNA. Much progress has been made on understanding and reverse-engineering TF network topologies using a range of experimental and theoretical methodologies. Less work has focused on using these models to examine how TF networks respond to changes in the cellular environment. METHODS: In this paper, we have developed a simple, pragmatic methodology, TIGERi (Transcription-factor-activity Illustrator for Global Explanation of Regulatory interaction), to model the response of an inferred TF network to changes in cellular environment. The methodology was tested using publicly available data comparing gene expression profiles of a mouse p38α (Mapk14) knock-out line to the original wild-type. RESULTS: Using the model, we have examined changes in the TF network resulting from the presence or absence of p38α. A part of this network was confirmed by experimental work in the original paper. Additional relationships were identified by our analysis, for example between p38α and HNF3, and between p38α and SOX9, and these are strongly supported by published evidence. FXR and MYC were also discovered in our analysis as two novel links of p38α. To provide a computational methodology to the biomedical communities that has more user-friendly interface, we also developed a standalone GUI (graphical user interface) software for TIGERi and it is freely available at https://github.com/namshik/tigeri/ . CONCLUSIONS: We therefore believe that our computational approach can identify new members of networks and new interactions between members that are supported by published data but have not been integrated into the existing network models. Moreover, ones who want to analyze their own data with TIGERi could use the software without any command line experience. This work could therefore accelerate researches in transcriptional gene regulation in higher eukaryotes.

Deregulation of the FOXM1 target gene network and its coregulatory partners in oesophageal adenocarcinoma.

BACKGROUND: Survival rates for oesophageal adenocarcinoma (OAC) remain disappointingly poor and current conventional treatment modalities have minimal impact on long-term survival. This is partly due to a lack of understanding of the molecular changes that occur in this disease. Previous studies have indicated that the transcription factor FOXM1 is commonly upregulated in this cancer type but the impact of this overexpression on gene expression in the context of OAC is largely unknown. FOXM1 does not function alone but works alongside the antagonistically-functioning co-regulatory MMB and DREAM complexes. METHODS: To establish how FOXM1 affects gene expression in OAC we have identified the FOXM1 target gene network in OAC-derived cells using ChIP-seq and determined the expression of both its coregulatory partners and members of this target gene network in OAC by digital transcript counting using the Nanostring gene expression assay. RESULTS: We find co-upregulation of FOXM1 with its target gene network in OAC. Furthermore, we find changes in the expression of its coregulatory partners, including co-upregulation of LIN9 and, surprisingly, reduced expression of LIN54. Mechanistically, we identify LIN9 as the direct binding partner for FOXM1 in the MMB complex. In the context of OAC, both coregulator (eg LIN54) and target gene (eg UHRF1) expression levels are predictive of disease stage. CONCLUSIONS: Together our data demonstrate that there are global changes to the FOXM1 regulatory network in OAC and the expression of components of this network help predict cancer prognosis.