A systematic review of biologically-informed deep learning models for cancer: fundamental trends for encoding and interpreting oncology data.

BACKGROUND: There is an increasing interest in the use of Deep Learning (DL) based methods as a supporting analytical framework in oncology. However, most direct applications of DL will deliver models with limited transparency and explainability, which constrain their deployment in biomedical settings. METHODS: This systematic review discusses DL models used to support inference in cancer biology with a particular emphasis on multi-omics analysis. It focuses on how existing models address the need for better dialogue with prior knowledge, biological plausibility and interpretability, fundamental properties in the biomedical domain. For this, we retrieved and analyzed 42 studies focusing on emerging architectural and methodological advances, the encoding of biological domain knowledge and the integration of explainability methods. RESULTS: We discuss the recent evolutionary arch of DL models in the direction of integrating prior biological relational and network knowledge to support better generalisation (e.g. pathways or Protein-Protein-Interaction networks) and interpretability. This represents a fundamental functional shift towards models which can integrate mechanistic and statistical inference aspects. We introduce a concept of bio-centric interpretability and according to its taxonomy, we discuss representational methodologies for the integration of domain prior knowledge in such models. CONCLUSIONS: The paper provides a critical outlook into contemporary methods for explainability and interpretability used in DL for cancer. The analysis points in the direction of a convergence between encoding prior knowledge and improved interpretability. We introduce bio-centric interpretability which is an important step towards formalisation of biological interpretability of DL models and developing methods that are less problem- or application-specific.

Methods for the Analysis of Topologically Associating Domains (TADs).

Chromatin folding in the 3D space of the nucleus can be explored through high-throughput chromosome conformation capture (Hi-C) approaches. These experiments quantify the number of interactions between any pair of genomic loci in the genome and, thus, allow building genome-scale maps of intra- and inter-chromosomal contacts (contact maps). Statistical and algorithmic analyses of Hi-C data consist in extracting information from these contact maps. One of the most striking patterns observed in intra-chromosomal Hi-C contact maps emerged from genomic regions that exhibit dense intra-region but sparse inter-region contacts. These have been termed topologically associating domains (TADs). The identification of TADs from Hi-C contact maps is of great interest as they have been shown to act as unit of chromosome organization and, potentially, functional activity. Several approaches have been developed to identify TADs (TAD callers). However, results from these methods are often dependent on data resolution and poorly concordant. In this chapter, we present four TAD callers and we provide detailed protocols for their use. In addition, we show how to compare TADs identified by different callers and how to assess the enrichment for TAD-associated biological features. TAD calling has become a key step in the study of chromatin 3D organization in different cellular contexts. Here we provide guidelines to improve the robustness and quality of these analyses.

Systematic assessment of gene co-regulation within chromatin domains determines differentially active domains across human cancers.

BACKGROUND: Spatial interactions and insulation of chromatin regions are associated with transcriptional regulation. Domains of frequent chromatin contacts are proposed as functional units, favoring and delimiting gene regulatory interactions. However, contrasting evidence supports the association between chromatin domains and transcription. RESULT: Here, we assess gene co-regulation in chromatin domains across multiple human cancers, which exhibit great transcriptional heterogeneity. Across all datasets, gene co-regulation is observed only within a small yet significant number of chromatin domains. We design an algorithmic approach to identify differentially active domains (DADo) between two conditions and show that these provide complementary information to differentially expressed genes. Domains comprising co-regulated genes are enriched in the less active B sub-compartments and for genes with similar function. Notably, differential activation of chromatin domains is not associated with major changes of domain boundaries, but rather with changes of sub-compartments and intra-domain contacts. CONCLUSION: Overall, gene co-regulation is observed only in a minority of chromatin domains, whose systematic identification will help unravel the relationship between chromatin structure and transcription.

Systematic inference and comparison of multi-scale chromatin sub-compartments connects spatial organization to cell phenotypes.

Chromatin compartmentalization reflects biological activity. However, inference of chromatin sub-compartments and compartment domains from chromosome conformation capture (Hi-C) experiments is limited by data resolution. As a result, these have been characterized only in a few cell types and systematic comparisons across multiple tissues and conditions are missing. Here, we present Calder, an algorithmic approach that enables the identification of multi-scale sub-compartments at variable data resolution. Calder allows to infer and compare chromatin sub-compartments and compartment domains in >100 cell lines. Our results reveal sub-compartments enriched for poised chromatin states and undergoing spatial repositioning during lineage differentiation and oncogenic transformation.

EZH2 oncogenic mutations drive epigenetic, transcriptional, and structural changes within chromatin domains.

Chromatin is organized into topologically associating domains (TADs) enriched in distinct histone marks. In cancer, gain-of-function mutations in the gene encoding the enhancer of zeste homolog 2 protein (EZH2) lead to a genome-wide increase in histone-3 Lys27 trimethylation (H3K27me3) associated with transcriptional repression. However, the effects of these epigenetic changes on the structure and function of chromatin domains have not been explored. Here, we found a functional interplay between TADs and epigenetic and transcriptional changes mediated by mutated EZH2. Altered EZH2 (p.Tyr646* (EZH2Y646X)) led to silencing of entire domains, synergistically inactivating multiple tumor suppressors. Intra-TAD gene silencing was coupled with changes of interactions between gene promoter regions. Notably, gene expression and chromatin interactions were restored by pharmacological inhibition of EZH2Y646X. Our results indicate that EZH2Y646X alters the topology and function of chromatin domains to promote synergistic oncogenic programs.

Comparison of computational methods for the identification of topologically associating domains.

BACKGROUND: Chromatin folding gives rise to structural elements among which are clusters of densely interacting DNA regions termed topologically associating domains (TADs). TADs have been characterized across multiple species, tissue types, and differentiation stages, sometimes in association with regulation of biological functions. The reliability and reproducibility of these findings are intrinsically related with the correct identification of these domains from high-throughput chromatin conformation capture (Hi-C) experiments. RESULTS: Here, we test and compare 22 computational methods to identify TADs across 20 different conditions. We find that TAD sizes and numbers vary significantly among callers and data resolutions, challenging the definition of an average TAD size, but strengthening the hypothesis that TADs are hierarchically organized domains, rather than disjoint structural elements. Performances of these methods differ based on data resolution and normalization strategy, but a core set of TAD callers consistently retrieve reproducible domains, even at low sequencing depths, that are enriched for TAD-associated biological features. CONCLUSIONS: This study provides a reference for the analysis of chromatin domains from Hi-C experiments and useful guidelines for choosing a suitable approach based on the experimental design, available data, and biological question of interest.