ExplaiNN: interpretable and transparent neural networks for genomics

Improving the performance of supervised deep learning for regulatory genomics using phylogenetic augmentation

MOTIVATION

Supervised deep learning is used to model the complex relationship between genomic sequence and regulatory function. Understanding how these models make predictions can provide biological insight into regulatory functions. Given the complexity of the sequence to regulatory function mapping (the cis-regulatory code), it has been suggested that the genome contains insufficient sequence variation to train models with suitable complexity. Data augmentation is a widely used approach to increase the data variation available for model training, however current data augmentation methods for genomic sequence data are limited.

RESULTS

Inspired by the success of comparative genomics, we show that augmenting genomic sequences with evolutionarily related sequences from other species, which we term phylogenetic augmentation, improves the performance of deep learning models trained on regulatory genomic sequences to predict high-throughput functional assay measurements. Additionally, we show that phylogenetic augmentation can rescue model performance when the training set is down-sampled and permits deep learning on a real-world small dataset, demonstrating that this approach improves data efficiency. Overall, this data augmentation method represents a solution for improving model performance that is applicable to many supervised deep-learning problems in genomics.

AVAILABILITY AND IMPLEMENTATION

The open-source GitHub repository agduncan94/phylogenetic_augmentation_paper includes the code for rerunning the analyses here and recreating the figures.

HOCOMOCO in 2024: a rebuild of the curated collection of binding models for human and mouse transcription factors

We present a major update of the HOCOMOCO collection that provides DNA binding specificity patterns of 949 human transcription factors and 720 mouse orthologs. To make this release, we performed motif discovery in peak sets that originated from 14 183 ChIP-Seq experiments and reads from 2554 HT-SELEX experiments yielding more than 400 thousand candidate motifs. The candidate motifs were annotated according to their similarity to known motifs and the hierarchy of DNA-binding domains of the respective transcription factors. Next, the motifs underwent human expert curation to stratify distinct motif subtypes and remove non-informative patterns and common artifacts. Finally, the curated subset of 100 thousand motifs was supplied to the automated benchmarking to select the best-performing motifs for each transcription factor. The resulting HOCOMOCO v12 core collection contains 1443 verified position weight matrices, including distinct subtypes of DNA binding motifs for particular transcription factors. In addition to the core collection, HOCOMOCO v12 provides motif sets optimized for the recognition of binding sites in vivo and in vitro, and for annotation of regulatory sequence variants. HOCOMOCO is available at https://hocomoco12.autosome.org and https://hocomoco.autosome.org.

JASPAR 2024: 20th anniversary of the open-access database of transcription factor binding profiles

JASPAR (https://jaspar.elixir.no/) is a widely-used open-access database presenting manually curated high-quality and non-redundant DNA-binding profiles for transcription factors (TFs) across taxa. In this 10th release and 20th-anniversary update, the CORE collection has expanded with 329 new profiles. We updated three existing profiles and provided orthogonal support for 72 profiles from the previous release's UNVALIDATED collection. Altogether, the JASPAR 2024 update provides a 20% increase in CORE profiles from the previous release. A trimming algorithm enhanced profiles by removing low information content flanking base pairs, which were likely uninformative (within the capacity of the PFM models) for TFBS predictions and modelling TF-DNA interactions. This release includes enhanced metadata, featuring a refined classification for plant TFs' structural DNA-binding domains. The new JASPAR collections prompt updates to the genomic tracks of predicted TF binding sites (TFBSs) in 8 organisms, with human and mouse tracks available as native tracks in the UCSC Genome browser. All data are available through the JASPAR web interface and programmatically through its API and the updated Bioconductor and pyJASPAR packages. Finally, a new TFBS extraction tool enables users to retrieve predicted JASPAR TFBSs intersecting their genomic regions of interest.

A fast machine learning dataloader for epigenetic tracks from BigWig files

SUMMARY

We created bigwig-loader, a data-loader for epigenetic profiles from BigWig files that decompresses and processes information for multiple intervals from multiple BigWig files in parallel. This is an access pattern needed to create training batches for typical machine learning models on epigenetics data. Using a new codec, the decompression can be done on a graphical processing unit (GPU) making it fast enough to create the training batches during training, mitigating the need for saving preprocessed training examples to disk.

AVAILABILITY AND IMPLEMENTATION

The bigwig-loader installation instructions and source code can be accessed at https://github.com/pfizer-opensource/bigwig-loader.

Transcriptional reprogramming by mutated IRF4 in lymphoma

Disease-causing mutations in genes encoding transcription factors (TFs) can affect TF interactions with their cognate DNA-binding motifs. Whether and how TF mutations impact upon the binding to TF composite elements (CE) and the interaction with other TFs is unclear. Here, we report a distinct mechanism of TF alteration in human lymphomas with perturbed B cell identity, in particular classic Hodgkin lymphoma. It is caused by a recurrent somatic missense mutation c.295 T > C (p.Cys99Arg; p.C99R) targeting the center of the DNA-binding domain of Interferon Regulatory Factor 4 (IRF4), a key TF in immune cells. IRF4-C99R fundamentally alters IRF4 DNA-binding, with loss-of-binding to canonical IRF motifs and neomorphic gain-of-binding to canonical and non-canonical IRF CEs. IRF4-C99R thoroughly modifies IRF4 function by blocking IRF4-dependent plasma cell induction, and up-regulates disease-specific genes in a non-canonical Activator Protein-1 (AP-1)-IRF-CE (AICE)-dependent manner. Our data explain how a single mutation causes a complex switch of TF specificity and gene regulation and open the perspective to specifically block the neomorphic DNA-binding activities of a mutant TF.

Human inborn errors of immunity associated with IRF4

The transcription factor interferon regulatory factor 4 (IRF4) belongs to the IRF family and has several important functions for the adaptive immune response. Mutations affecting IRF family members IRF1, IRF3, IRF7, IRF8, or IRF9 have been described in patients presenting with inborn errors of immunity (IEI) highlighting the importance of these factors for the cellular host defense against mycobacterial and/or viral infections. IRF4 deficiency and haploinsufficiency have been associated with IEI. More recently, two novel IRF4 disease-causing mechanisms have been described due to the characterization of IEI patients presenting with cellular immunodeficiency associated with agammaglobulinemia. Here, we review the phenotypes and physiopathological mechanisms underlying IEI of IRF family members and, in particular, IRF4.

An intrinsically interpretable neural network architecture for sequence-to-function learning

MOTIVATION

Sequence-based deep learning approaches have been shown to predict a multitude of functional genomic readouts, including regions of open chromatin and RNA expression of genes. However, a major limitation of current methods is that model interpretation relies on computationally demanding post hoc analyses, and even then, one can often not explain the internal mechanics of highly parameterized models. Here, we introduce a deep learning architecture called totally interpretable sequence-to-function model (tiSFM). tiSFM improves upon the performance of standard multilayer convolutional models while using fewer parameters. Additionally, while tiSFM is itself technically a multilayer neural network, internal model parameters are intrinsically interpretable in terms of relevant sequence motifs.

RESULTS

We analyze published open chromatin measurements across hematopoietic lineage cell-types and demonstrate that tiSFM outperforms a state-of-the-art convolutional neural network model custom-tailored to this dataset. We also show that it correctly identifies context-specific activities of transcription factors with known roles in hematopoietic differentiation, including Pax5 and Ebf1 for B-cells, and Rorc for innate lymphoid cells. tiSFM's model parameters have biologically meaningful interpretations, and we show the utility of our approach on a complex task of predicting the change in epigenetic state as a function of developmental transition.

AVAILABILITY AND IMPLEMENTATION

The source code, including scripts for the analysis of key findings, can be found at https://github.com/boooooogey/ATAConv, implemented in Python.