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.

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.

Refining the genomic determinants underlying escape from X-chromosome inactivation.

X-chromosome inactivation (XCI) epigenetically silences one X chromosome in every cell in female mammals. Although the majority of X-linked genes are silenced, in humans 20% or more are able to escape inactivation and continue to be expressed. Such escape genes are important contributors to sex differences in gene expression, and may impact the phenotypes of X aneuploidies; yet the mechanisms regulating escape from XCI are not understood. We have performed an enrichment analysis of transcription factor binding on the X chromosome, providing new evidence for enriched factors at the transcription start sites of escape genes. The top escape-enriched transcription factors were detected at the RPS4X promoter, a well-described human escape gene previously demonstrated to escape from XCI in a transgenic mouse model. Using a cell line model system that allows for targeted integration and inactivation of transgenes on the mouse X chromosome, we further assessed combinations of RPS4X promoter and genic elements for their ability to drive escape from XCI. We identified a small transgenic construct of only 6 kb capable of robust escape from XCI, establishing that gene-proximal elements are sufficient to permit escape, and highlighting the additive effect of multiple elements that work together in a context-specific fashion.

ExplaiNN: interpretable and transparent neural networks for genomics.

Deep learning models such as convolutional neural networks (CNNs) excel in genomic tasks but lack interpretability. We introduce ExplaiNN, which combines the expressiveness of CNNs with the interpretability of linear models. ExplaiNN can predict TF binding, chromatin accessibility, and de novo motifs, achieving performance comparable to state-of-the-art methods. Its predictions are transparent, providing global (cell state level) as well as local (individual sequence level) biological insights into the data. ExplaiNN can serve as a plug-and-play platform for pretrained models and annotated position weight matrices. ExplaiNN aims to accelerate the adoption of deep learning in genomic sequence analysis by domain experts.

OnTarget: in silico design of MiniPromoters for targeted delivery of expression.

MiniPromoters, or compact promoters, are short DNA sequences that can drive expression in specific cells and tissues. While broadly useful, they are of high relevance to gene therapy due to their role in enabling precise control of where a therapeutic gene will be expressed. Here, we present OnTarget (http://ontarget.cmmt.ubc.ca), a webserver that streamlines the MiniPromoter design process. Users only need to specify a gene of interest or custom genomic coordinates on which to focus the identification of promoters and enhancers, and can also provide relevant cell-type-specific genomic evidence (e.g. accessible chromatin regions, histone modifications, etc.). OnTarget combines the provided data with internal data to identify candidate promoters and enhancers and design MiniPromoters. To illustrate the utility of OnTarget, we designed and characterized two MiniPromoters targeting different cell populations relevant to Parkinson Disease.

In silico discovery of small molecules for efficient stem cell differentiation into definitive endoderm.

Improving methods for human embryonic stem cell differentiation represents a challenge in modern regenerative medicine research. Using drug repurposing approaches, we discover small molecules that regulate the formation of definitive endoderm. Among them are inhibitors of known processes involved in endoderm differentiation (mTOR, PI3K, and JNK pathways) and a new compound, with an unknown mechanism of action, capable of inducing endoderm formation in the absence of growth factors in the media. Optimization of the classical protocol by inclusion of this compound achieves the same differentiation efficiency with a 90% cost reduction. The presented in silico procedure for candidate molecule selection has broad potential for improving stem cell differentiation protocols.

Prediction of Blood Risk Score in Diabetes Using Deep Neural Networks.

Improving the prediction of blood glucose concentration may improve the quality of life of people living with type 1 diabetes by enabling them to better manage their care. Given the anticipated benefits of such a prediction, numerous methods have been proposed. Rather than attempting to predict glucose concentration, a deep learning framework for prediction is proposed in which prediction is performed using a scale for hypo- and hyper-glycemia risk. Using the blood glucose risk score formula proposed by Kovatchev et al., models with different architectures were trained, including, a recurrent neural network (RNN), a gated recurrent unit (GRU), a long short-term memory (LSTM) network, and an encoder-like convolutional neural network (CNN). The models were trained using the OpenAPS Data Commons data set, comprising 139 individuals, each with tens of thousands of continuous glucose monitor (CGM) data points. The training set was composed of 7% of the data set, while the remaining was used for testing. Performance comparisons between the different architectures are presented and discussed. To evaluate these predictions, performance results are compared with the last measurement (LM) prediction, through a sample-and-hold approach continuing the last known measurement forward. The results obtained are competitive when compared to other deep learning methods. A root mean squared error (RMSE) of 16 mg/dL, 24 mg/dL, and 37 mg/dL were obtained for CNN prediction horizons of 15, 30, and 60 min, respectively. However, no significant improvements were found for the deep learning models compared to LM prediction. Performance was found to be highly dependent on architecture and the prediction horizon. Lastly, a metric to assess model performance by weighing each prediction point error with the corresponding blood glucose risk score is proposed. Two main conclusions are drawn. Firstly, going forward, there is a need to benchmark model performance using LM prediction to enable the comparison between results obtained from different data sets. Secondly, model-agnostic data-driven deep learning models may only be meaningful when combined with mechanistic physiological models; here, it is argued that neural ordinary differential equations may combine the best of both approaches. These findings are based on the OpenAPS Data Commons data set and are to be validated in other independent data sets.

Obtaining genetics insights from deep learning via explainable artificial intelligence.

Artificial intelligence (AI) models based on deep learning now represent the state of the art for making functional predictions in genomics research. However, the underlying basis on which predictive models make such predictions is often unknown. For genomics researchers, this missing explanatory information would frequently be of greater value than the predictions themselves, as it can enable new insights into genetic processes. We review progress in the emerging area of explainable AI (xAI), a field with the potential to empower life science researchers to gain mechanistic insights into complex deep learning models. We discuss and categorize approaches for model interpretation, including an intuitive understanding of how each approach works and their underlying assumptions and limitations in the context of typical high-throughput biological datasets.

Human complete NFAT1 deficiency causes a triad of joint contractures, osteochondromas, and B-cell malignancy.

The discovery of humans with monogenic disorders has a rich history of generating new insights into biology. Here we report the first human identified with complete deficiency of nuclear factor of activated T cells 1 (NFAT1). NFAT1, encoded by NFATC2, mediates calcium-calcineurin signals that drive cell activation, proliferation, and survival. The patient is homozygous for a damaging germline NFATC2 variant (c.2023_2026delTACC; p.Tyr675Thrfs∗18) and presented with joint contractures, osteochondromas, and recurrent B-cell lymphoma. Absence of NFAT1 protein in chondrocytes caused enrichment in prosurvival and inflammatory genes. Systematic single-cell-omic analyses in PBMCs revealed an environment that promotes lymphomagenesis with accumulation of naïve B cells (enriched for oncogenic signatures MYC and JAK1), exhausted CD4+ T cells, impaired T follicular helper cells, and aberrant CD8+ T cells. This work highlights the pleiotropic role of human NFAT1, will empower the diagnosis of additional patients with NFAT1 deficiency, and further defines the detrimental effects associated with long-term use of calcineurin inhibitors.

RevUP: an online scoring system for regulatory variants implicated in rare diseases.

SUMMARY: To address the difficulty in assessing the implication of regulatory variants in diseases, a scoring scheme previously published allows the calculation of the Regulatory Variant Evidence score (RVE-score). The score represents the accumulated evidence for a causative role of a regulatory variant in a disease. Regulatory Evidence for Variants Underlying Phenotypes was built to calculate the RVE-score of regulatory variants, based on the 24 criteria, with a hybrid approach combining information retrieved from public databases and user input. AVAILABILITY AND IMPLEMENTATION: RevUP is freely available at http://www.revup-classifier.ca. The source code is available at https://github.com/wassermanlab/revup. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Linkage analysis identifies an isolated strabismus locus at 14q12 overlapping with FOXG1 syndrome region.

Strabismus is a common condition, affecting 1%-4% of individuals. Isolated strabismus has been studied in families with Mendelian inheritance patterns. Despite the identification of multiple loci via linkage analyses, no specific genes have been identified from these studies. The current study is based on a seven-generation family with isolated strabismus inherited in an autosomal dominant manner. A total of 13 individuals from a common ancestor have been included for linkage analysis. Among these, nine are affected and four are unaffected. A single linkage signal has been identified at an 8.5 Mb region of chromosome 14q12 with a multipoint LOD (logarithm of the odds) score of 4.69. Disruption of this locus is known to cause FOXG1 syndrome (or congenital Rett syndrome; OMIM #613454 and *164874), in which 84% of affected individuals present with strabismus. With the incorporation of next-generation sequencing and in-depth bioinformatic analyses, a 4 bp non-coding deletion was prioritised as the top candidate for the observed strabismus phenotype. The deletion is predicted to disrupt regulation of FOXG1, which encodes a transcription factor of the Forkhead family. Suggestive of an autoregulation effect, the disrupted sequence matches the consensus FOXG1 and Forkhead family transcription factor binding site and has been observed in previous ChIP-seq studies to be bound by Foxg1 in early mouse brain development. Future study of this specific deletion may shed light on the regulation of FOXG1 expression and may enhance our understanding of the mechanisms contributing to strabismus and FOXG1 syndrome.

JASPAR 2022: the 9th release of the open-access database of transcription factor binding profiles.

JASPAR (http://jaspar.genereg.net/) is an open-access database containing manually curated, non-redundant transcription factor (TF) binding profiles for TFs across six taxonomic groups. In this 9th release, we expanded the CORE collection with 341 new profiles (148 for plants, 101 for vertebrates, 85 for urochordates, and 7 for insects), which corresponds to a 19% expansion over the previous release. We added 298 new profiles to the Unvalidated collection when no orthogonal evidence was found in the literature. All the profiles were clustered to provide familial binding profiles for each taxonomic group. Moreover, we revised the structural classification of DNA binding domains to consider plant-specific TFs. This release introduces word clouds to represent the scientific knowledge associated with each TF. We updated the genome tracks of TFBSs predicted with JASPAR profiles in eight organisms; the human and mouse TFBS predictions can be visualized as native tracks in the UCSC Genome Browser. Finally, we provide a new tool to perform JASPAR TFBS enrichment analysis in user-provided genomic regions. All the data is accessible through the JASPAR website, its associated RESTful API, the R/Bioconductor data package, and a new Python package, pyJASPAR, that facilitates serverless access to the data.

Biologically relevant transfer learning improves transcription factor binding prediction.

BACKGROUND: Deep learning has proven to be a powerful technique for transcription factor (TF) binding prediction but requires large training datasets. Transfer learning can reduce the amount of data required for deep learning, while improving overall model performance, compared to training a separate model for each new task. RESULTS: We assess a transfer learning strategy for TF binding prediction consisting of a pre-training step, wherein we train a multi-task model with multiple TFs, and a fine-tuning step, wherein we initialize single-task models for individual TFs with the weights learned by the multi-task model, after which the single-task models are trained at a lower learning rate. We corroborate that transfer learning improves model performance, especially if in the pre-training step the multi-task model is trained with biologically relevant TFs. We show the effectiveness of transfer learning for TFs with ~ 500 ChIP-seq peak regions. Using model interpretation techniques, we demonstrate that the features learned in the pre-training step are refined in the fine-tuning step to resemble the binding motif of the target TF (i.e., the recipient of transfer learning in the fine-tuning step). Moreover, pre-training with biologically relevant TFs allows single-task models in the fine-tuning step to learn useful features other than the motif of the target TF. CONCLUSIONS: Our results confirm that transfer learning is a powerful technique for TF binding prediction.

Secondary biogenic amine deficiencies: genetic etiology, therapeutic interventions, and clinical effects.

Monoamine neurotransmitter disorders present predominantly with neurologic features, including dystonic or dyskinetic cerebral palsy and movement disorders. Genetic conditions that lead to secondary defects in the synthesis, catabolism, transport, and metabolism of biogenic amines can lead to neurotransmitter abnormalities, which can present with similar features. Eleven patients with secondary neurotransmitter abnormalities were enrolled between 2011 and 2015. All patients underwent research-based whole exome and/or whole genome sequencing (WES/WGS). A trial of treatment with levodopa/carbidopa and 5-hydroxytryptophan was initiated. In six families with abnormal neurotransmitter profiles and neurological phenotypes, variants in known disease-causing genes (KCNJ6, SCN2A, CSTB in 2 siblings, NRNX1, KIF1A and PAK3) were identified, while one patient had a variant of uncertain significance in a candidate gene (DLG4) that may explain her phenotype. In 3 patients, no compelling candidate genes were identified. A trial of neurotransmitter replacement therapy led to improvement in motor and behavioral symptoms in all but two patients. The patient with KCNJ6 variant did not respond to L-dopa therapy, but rather experienced increased dyskinetic movements even at low dose of medication. The patient's symptoms harboring the NRNX1 deletion remained unaltered. This study demonstrates the utility of genome-wide sequencing in further understanding the etiology and pathophysiology of neurometabolic conditions, and the potential of secondary neurotransmitter deficiencies to serve as novel therapeutic targets. As there was a largely favorable response to therapy in our case series, a careful trial of neurotransmitter replacement therapy should be considered in patients with cerebrospinal fluid (CSF) monoamines below reference range.

Discovery of widespread transcription initiation at microsatellites predictable by sequence-based deep neural network.

Using the Cap Analysis of Gene Expression (CAGE) technology, the FANTOM5 consortium provided one of the most comprehensive maps of transcription start sites (TSSs) in several species. Strikingly, ~72% of them could not be assigned to a specific gene and initiate at unconventional regions, outside promoters or enhancers. Here, we probe these unassigned TSSs and show that, in all species studied, a significant fraction of CAGE peaks initiate at microsatellites, also called short tandem repeats (STRs). To confirm this transcription, we develop Cap Trap RNA-seq, a technology which combines cap trapping and long read MinION sequencing. We train sequence-based deep learning models able to predict CAGE signal at STRs with high accuracy. These models unveil the importance of STR surrounding sequences not only to distinguish STR classes, but also to predict the level of transcription initiation. Importantly, genetic variants linked to human diseases are preferentially found at STRs with high transcription initiation level, supporting the biological and clinical relevance of transcription initiation at STRs. Together, our results extend the repertoire of non-coding transcription associated with DNA tandem repeats and complexify STR polymorphism.

Adult GAMT deficiency: A literature review and report of two siblings.

Guanidinoacetate methyltransferase (GAMT) deficiency is a creatine deficiency disorder and an inborn error of metabolism presenting with progressive intellectual and neurological deterioration. As most cases are identified and treated in early childhood, adult phenotypes that can help in understanding the natural history of the disorder are rare. We describe two adult cases of GAMT deficiency from a consanguineous family in Pakistan that presented with a history of global developmental delay, cognitive impairments, excessive drooling, behavioral abnormalities, contractures and apparent bone deformities initially presumed to be the reason for abnormal gait. Exome sequencing identified a homozygous nonsense variant in GAMT: NM_000156.5:c.134G>A (p.Trp45*). We also performed a literature review and compiled the genetic and clinical characteristics of all adult cases of GAMT deficiency reported to date. When compared to the adult cases previously reported, the musculoskeletal phenotype and the rapidly progressive nature of neurological and motor decline seen in our patients is striking. This study presents an opportunity to gain insights into the adult presentation of GAMT deficiency and highlights the need for in-depth evaluation and reporting of clinical features to expand our understanding of the phenotypic spectrum.

Demonstrating the utility of flexible sequence queries against indexed short reads with FlexTyper.

Across the life sciences, processing next generation sequencing data commonly relies upon a computationally expensive process where reads are mapped onto a reference sequence. Prior to such processing, however, there is a vast amount of information that can be ascertained from the reads, potentially obviating the need for processing, or allowing optimized mapping approaches to be deployed. Here, we present a method termed FlexTyper which facilitates a "reverse mapping" approach in which high throughput sequence queries, in the form of k-mer searches, are run against indexed short-read datasets in order to extract useful information. This reverse mapping approach enables the rapid counting of target sequences of interest. We demonstrate FlexTyper's utility for recovering depth of coverage, and accurate genotyping of SNP sites across the human genome. We show that genotyping unmapped reads can correctly inform a sample's population, sex, and relatedness in a family setting. Detection of pathogen sequences within RNA-seq data was sensitive and accurate, performing comparably to existing methods, but with increased flexibility. We present two examples of ways in which this flexibility allows the analysis of genome features not well-represented in a linear reference. First, we analyze contigs from African genome sequencing studies, showing how they distribute across families from three distinct populations. Second, we show how gene-marking k-mers for the killer immune receptor locus allow allele detection in a region that is challenging for standard read mapping pipelines. The future adoption of the reverse mapping approach represented by FlexTyper will be enabled by more efficient methods for FM-index generation and biology-informed collections of reference queries. In the long-term, selection of population-specific references or weighting of edges in pan-population reference genome graphs will be possible using the FlexTyper approach. FlexTyper is available at https://github.com/wassermanlab/OpenFlexTyper.

Targeting AXL kinase sensitizes leukemic stem and progenitor cells to venetoclax treatment in acute myeloid leukemia.

The abundance of genetic abnormalities and phenotypic heterogeneities in acute myeloid leukemia (AML) poses significant challenges to the development of improved treatments. Here, we demonstrated that a key growth arrest-specific gene 6/AXL axis is highly activated in cells from patients with AML, particularly in stem/progenitor cells. We developed a potent selective AXL inhibitor that has favorable pharmaceutical properties and efficacy against preclinical patient-derived xenotransplantation (PDX) models of AML. Importantly, inhibition of AXL sensitized AML stem/progenitor cells to venetoclax treatment, with strong synergistic effects in vitro and in PDX models. Mechanistically, single-cell RNA-sequencing and functional validation studies uncovered that AXL inhibition, alone or in combination with venetoclax, potentially targets intrinsic metabolic vulnerabilities of AML stem/progenitor cells and shows a distinct transcriptomic profile and inhibits mitochondrial oxidative phosphorylation. Inhibition of AXL or BCL-2 also differentially targets key signaling proteins to synergize in leukemic cell killing. These findings have a direct translational impact on the treatment of AML and other cancers with high AXL activity.

Human MiniPromoters for ocular-rAAV expression in ON bipolar, cone, corneal, endothelial, Müller glial, and PAX6 cells.

Small and cell-type restricted promoters are important tools for basic and preclinical research, and clinical delivery of gene therapies. In clinical gene therapy, ophthalmic trials have been leading the field, with over 50% of ocular clinical trials using promoters that restrict expression based on cell type. Here, 19 human DNA MiniPromoters were bioinformatically designed for rAAV, tested by neonatal intravenous delivery in mouse, and successful MiniPromoters went on to be tested by intravitreal, subretinal, intrastromal, and/or intravenous delivery in adult mouse. We present promoter development as an overview for each cell type, but only show results in detail for the recommended MiniPromoters: Ple265 and Ple341 (PCP2) ON bipolar, Ple349 (PDE6H) cone, Ple253 (PITX3) corneal stroma, Ple32 (CLDN5) endothelial cells of the blood-retina barrier, Ple316 (NR2E1) Müller glia, and Ple331 (PAX6) PAX6 positive. Overall, we present a resource of new, redesigned, and improved MiniPromoters for ocular gene therapy that range in size from 784 to 2484 bp, and from weaker, equal, or stronger in strength relative to the ubiquitous control promoter smCBA. All MiniPromoters will be useful for therapies involving small regulatory RNA and DNA, and proteins ranging from 517 to 1084 amino acids, representing 62.9-90.2% of human proteins.