Deep learning sequence-based ab initio prediction of variant effects on expression and disease risk

DeepCORE: An interpretable multi-view deep neural network model to detect co-operative regulatory elements

Gene transcription is an essential process involved in all aspects of cellular functions with significant impact on biological traits and diseases. This process is tightly regulated by multiple elements that co-operate to jointly modulate the transcription levels of target genes. To decipher the complicated regulatory network, we present a novel multi-view attention-based deep neural network that models the relationship between genetic, epigenetic, and transcriptional patterns and identifies co-operative regulatory elements (COREs). We applied this new method, named DeepCORE, to predict transcriptomes in various tissues and cell lines, which outperformed the state-of-the-art algorithms. Furthermore, DeepCORE contains an interpreter that extracts the attention values embedded in the deep neural network, maps the attended regions to putative regulatory elements, and infers COREs based on correlated attentions. The identified COREs are significantly enriched with known promoters and enhancers. Novel regulatory elements discovered by DeepCORE showed epigenetic signatures consistent with the status of histone modification marks.

Big data and artificial intelligence-aided crop breeding: Progress and prospects

The past decade has witnessed rapid developments in gene discovery, biological big data (BBD), artificial intelligence (AI)-aided technologies, and molecular breeding. These advancements are expected to accelerate crop breeding under the pressure of increasing demands for food. Here, we first summarize current breeding methods and discuss the need for new ways to support breeding efforts. Then, we review how to combine BBD and AI technologies for genetic dissection, exploring functional genes, predicting regulatory elements and functional domains, and phenotypic prediction. Finally, we propose the concept of intelligent precision design breeding (IPDB) driven by AI technology and offer ideas about how to implement IPDB. We hope that IPDB will enhance the predictability, efficiency, and cost of crop breeding compared with current technologies. As an example of IPDB, we explore the possibilities offered by CropGPT, which combines biological techniques, bioinformatics, and breeding art from breeders, and presents an open, shareable, and cooperative breeding system. IPDB provides integrated services and communication platforms for biologists, bioinformatics experts, germplasm resource specialists, breeders, dealers, and farmers, and should be well suited for future breeding.

Training deep learning models on personalized genomic sequences improves variant effect prediction

Sequence-to-function models have broad applications in interpreting the molecular impact of genetic variation, yet have been criticized for poor performance in this task. Here we show that training models on functional genomic data with matched personal genomes improves their performance at variant effect prediction. Variant effect representations are retained even when transfer learning models to unseen cellular contexts and experimental readouts. Our results have implications for interpreting trait-associated genetic variation.

A novel interpretable deep learning-based computational framework designed synthetic enhancers with broad cross-species activity

Enhancers play a critical role in dynamically regulating spatial-temporal gene expression and establishing cell identity, underscoring the significance of designing them with specific properties for applications in biosynthetic engineering and gene therapy. Despite numerous high-throughput methods facilitating genome-wide enhancer identification, deciphering the sequence determinants of their activity remains challenging. Here, we present the DREAM (DNA cis-Regulatory Elements with controllable Activity design platforM) framework, a novel deep learning-based approach for synthetic enhancer design. Proficient in uncovering subtle and intricate patterns within extensive enhancer screening data, DREAM achieves cutting-edge sequence-based enhancer activity prediction and highlights critical sequence features implicating strong enhancer activity. Leveraging DREAM, we have engineered enhancers that surpass the potency of the strongest enhancer within the Drosophila genome by approximately 3.6-fold. Remarkably, these synthetic enhancers exhibited conserved functionality across species that have diverged more than billion years, indicating that DREAM was able to learn highly conserved enhancer regulatory grammar. Additionally, we designed silencers and cell line-specific enhancers using DREAM, demonstrating its versatility. Overall, our study not only introduces an interpretable approach for enhancer design but also lays out a general framework applicable to the design of other types of cis-regulatory elements.

Enhancing personalized gene expression prediction from DNA sequences using genomic foundation models

Artificial intelligence (AI)/deep learning (DL) models that predict molecular phenotypes like gene expression directly from DNA sequences have recently emerged. While these models have proven effective at capturing the variation across genes, their ability to explain inter-individual differences has been limited. We hypothesize that the performance gap can be narrowed through the use of pre-trained embeddings from the Nucleotide Transformer, a large foundation model trained on 3,000+ genomes. We train a transformer model using the pre-trained embeddings and compare its predictive performance to Enformer, the current state-of-the-art model, using genotype and expression data from 290 individuals. Our model significantly outperforms Enformer in terms of correlation across individuals, and narrows the performance gap with an elastic net regression approach that uses just the genetic variants as predictors. Although simple regression models have their advantages in personalized prediction tasks, DL approaches based on foundation models pre-trained on diverse genomes have unique strengths in flexibility and interpretability. With further methodological and computational improvements with more training data, these models may eventually predict molecular phenotypes from DNA sequences with an accuracy surpassing that of regression-based approaches. Our work demonstrates the potential for large pre-trained AI/DL models to advance functional genomics.

Fundamentals for predicting transcriptional regulations from DNA sequence patterns

Cell-type-specific regulatory elements, cataloged through extensive experiments and bioinformatics in large-scale consortiums, have enabled enrichment analyses of genetic associations that primarily utilize positional information of the regulatory elements. These analyses have identified cell types and pathways genetically associated with human complex traits. However, our understanding of detailed allelic effects on these elements' activities and on-off states remains incomplete, hampering the interpretation of human genetic study results. This review introduces machine learning methods to learn sequence-dependent transcriptional regulation mechanisms from DNA sequences for predicting such allelic effects (not associations). We provide a concise history of machine-learning-based approaches, the requirements, and the key computational processes, focusing on primers in machine learning. Convolution and self-attention, pivotal in modern deep-learning models, are explained through geometrical interpretations using dot products. This facilitates understanding of the concept and why these have been used for machine learning for DNA sequences. These will inspire further research in this genetics and genomics field.

Genotype imputation methods for whole and complex genomic regions utilizing deep learning technology

The imputation of unmeasured genotypes is essential in human genetic research, particularly in enhancing the power of genome-wide association studies and conducting subsequent fine-mapping. Recently, several deep learning-based genotype imputation methods for genome-wide variants with the capability of learning complex linkage disequilibrium patterns have been developed. Additionally, deep learning-based imputation has been applied to a distinct genomic region known as the major histocompatibility complex, referred to as HLA imputation. Despite their various advantages, the current deep learning-based genotype imputation methods do have certain limitations and have not yet become standard. These limitations include the modest accuracy improvement over statistical and conventional machine learning-based methods. However, their benefits include other aspects, such as their "reference-free" nature, which ensures complete privacy protection, and their higher computational efficiency. Furthermore, the continuing evolution of deep learning technologies is expected to contribute to further improvements in prediction accuracy and usability in the future.

Leveraging State-of-the-Art AI Algorithms in Personalized Oncology: From Transcriptomics to Treatment

BACKGROUND

Continuous breakthroughs in computational algorithms have positioned AI-based models as some of the most sophisticated technologies in the healthcare system. AI shows dynamic contributions in advancing various medical fields involving data interpretation and monitoring, imaging screening and diagnosis, and treatment response and survival prediction. Despite advances in clinical oncology, more effort must be employed to tailor therapeutic plans based on each patient's unique transcriptomic profile within the precision/personalized oncology frame. Furthermore, the standard analysis method is not compatible with the comprehensive deciphering of significant data streams, thus precluding the prediction of accurate treatment options.

METHODOLOGY

We proposed a novel approach that includes obtaining different tumour tissues and preparing RNA samples for comprehensive transcriptomic interpretation using specifically trained, programmed, and optimized AI-based models for extracting large data volumes, refining, and analyzing them. Next, the transcriptomic results will be scanned against an expansive drug library to predict the response of each target to the tested drugs. The obtained target-drug combination/s will be then validated using in vitro and in vivo experimental models. Finally, the best treatment combination option/s will be introduced to the patient. We also provided a comprehensive review discussing AI models' recent innovations and implementations to aid in molecular diagnosis and treatment planning.

RESULTS

The expected transcriptomic analysis generated by the AI-based algorithms will provide an inclusive genomic profile for each patient, containing statistical and bioinformatics analyses, identification of the dysregulated pathways, detection of the targeted genes, and recognition of molecular biomarkers. Subjecting these results to the prediction and pairing AI-based processes will result in statistical graphs presenting each target's likely response rate to various treatment options. Different in vitro and in vivo investigations will further validate the selection of the target drug/s pairs.

CONCLUSIONS

Leveraging AI models will provide more rigorous manipulation of large-scale datasets on specific cancer care paths. Such a strategy would shape treatment according to each patient's demand, thus fortifying the avenue of personalized/precision medicine. Undoubtedly, this will assist in improving the oncology domain and alleviate the burden of clinicians in the coming decade.

Integrative identification of non-coding regulatory regions driving metastatic prostate cancer

Large-scale sequencing efforts have been undertaken to understand the mutational landscape of the coding genome. However, the vast majority of variants occur within non-coding genomic regions. We designed an integrative computational and experimental framework to identify recurrently mutated non-coding regulatory regions that drive tumor progression. Applying this framework to sequencing data from a large prostate cancer patient cohort revealed a large set of candidate drivers. We used (1) in silico analyses, (2) massively parallel reporter assays, and (3) in vivo CRISPR interference screens to systematically validate metastatic castration-resistant prostate cancer (mCRPC) drivers. One identified enhancer region, GH22I030351, acts on a bidirectional promoter to simultaneously modulate expression of the U2-associated splicing factor SF3A1 and chromosomal protein CCDC157. SF3A1 and CCDC157 promote tumor growth in vivo. We nominated a number of transcription factors, notably SOX6, to regulate expression of SF3A1 and CCDC157. Our integrative approach enables the systematic detection of non-coding regulatory regions that drive human cancers.

MLSNet: a deep learning model for predicting transcription factor binding sites

Accurate prediction of transcription factor binding sites (TFBSs) is essential for understanding gene regulation mechanisms and the etiology of diseases. Despite numerous advances in deep learning for predicting TFBSs, their performance can still be enhanced. In this study, we propose MLSNet, a novel deep learning architecture designed specifically to predict TFBSs. MLSNet innovatively integrates multisize convolutional fusion with long short-term memory (LSTM) networks to effectively capture DNA-sparse higher-order sequence features. Further, MLSNet incorporates super token attention and Bi-LSTM to systematically extract and integrate higher-order DNA shape features. Experimental results on 165 ChIP-seq (chromatin immunoprecipitation followed by sequencing) datasets indicate that MLSNet consistently outperforms several state-of-the-art algorithms in the prediction of TFBSs. Specifically, MLSNet reports average metrics: 0.8306 for ACC, 0.8992 for AUROC, and 0.9035 for AUPRC, surpassing the second-best methods by 1.82%, 1.68%, and 1.54%, respectively. This research delineates the effectiveness of combining multi-size convolutional layers with LSTM and DNA shape-based features in enhancing predictive accuracy. Moreover, this study comprehensively assesses the variability in model performance across different cell lines and transcription factors. The source code of MLSNet is available at https://github.com/minghaidea/MLSNet.

Quick and effective approximation of in silico saturation mutagenesis experiments with first-order taylor expansion

To understand the decision process of genomic sequence-to-function models, explainable AI algorithms determine the importance of each nucleotide in a given input sequence to the model's predictions and enable discovery of cis-regulatory motifs for gene regulation. The most commonly applied method is in silico saturation mutagenesis (ISM) because its per-nucleotide importance scores can be intuitively understood as the computational counterpart to in vivo saturation mutagenesis experiments. While ISM is highly interpretable, it is computationally challenging to perform for many sequences, and becomes prohibitive as the length of the input sequences and size of the model grows. Here, we use the first-order Taylor approximation to approximate ISM values from the model's gradient, which reduces its computation cost to a single forward pass for an input sequence. We show that the Taylor ISM (TISM) approximation is robust across different model ablations, random initializations, training parameters, and dataset sizes.

ProSol-multi: Protein solubility prediction via amino acids multi-level correlation and discriminative distribution

Protein solubility prediction is useful for the careful selection of highly effective candidate proteins for drug development. In recombinant proteins synthesis, solubility prediction is valuable for optimizing key protein characteristics, including stability, functionality, and ease of purification. It contains valuable information about potential biomarkers or therapeutic targets and helps in early forecasting of neurodegenerative diseases, cancer, and cardiovascular disorders. Traditional wet-lab experimental protein solubility prediction approaches are error-prone, time-consuming, and costly. Researchers harnessed the competence of Artificial Intelligence approaches for replacing experimental approaches with computational predictors. These predictors inferred the solubility of proteins by analyzing amino acids distributions in raw protein sequences. There is still a lot of room for the development of robust computational predictors because existing predictors remain fail in extracting comprehensive discriminative distribution of amino acids. To more precisely discriminate soluble proteins from insoluble proteins, this paper presents ProSol-Multi predictor that makes use of a novel MLCDE encoder and Random Forest classifier. MLCDE encoder transforms protein sequences into informative statistical vectors by capturing amino acids multi-level correlation and discriminative distribution within raw protein sequences. The performance of proposed encoder is evaluated against 56 existing protein sequence encoding methods on a widely used protein solubility prediction benchmark dataset under two different experimental settings namely intrinsic and extrinsic. Intrinsic evaluation reveals that from all sequence encoders, proposed MLCDE encoder manages to generate non-overlapping clusters of soluble and insoluble classes. In extrinsic evaluation, 10 machine learning classifiers achieve better performance with proposed MLCDE encoder as compared to 56 existing protein sequence encoders. Moreover, across 4 public benchmark datasets, proposed ProSol-Multi predictor outshines 20 existing predictors by an average accuracy of 3%, MCC and AU-ROC of 2%. ProSol-Multi interactive web application is available at https://sds_genetic_analysis.opendfki.de/ProSol-Multi.

DeepCBA: A deep learning framework for gene expression prediction in maize based on DNA sequences and chromatin interactions

Chromatin interactions create spatial proximity between distal regulatory elements and target genes in the genome, which has an important impact on gene expression, transcriptional regulation, and phenotypic traits. To date, several methods have been developed for predicting gene expression. However, existing methods do not take into consideration the effect of chromatin interactions on target gene expression, thus potentially reducing the accuracy of gene expression prediction and mining of important regulatory elements. In this study, we developed a highly accurate deep learning-based gene expression prediction model (DeepCBA) based on maize chromatin interaction data. Compared with existing models, DeepCBA exhibits higher accuracy in expression classification and expression value prediction. The average Pearson correlation coefficients (PCCs) for predicting gene expression using gene promoter proximal interactions, proximal-distal interactions, and both proximal and distal interactions were 0.818, 0.625, and 0.929, respectively, representing an increase of 0.357, 0.16, and 0.469 over the PCCs obtained with traditional methods that use only gene proximal sequences. Some important motifs were identified through DeepCBA; they were enriched in open chromatin regions and expression quantitative trait loci and showed clear tissue specificity. Importantly, experimental results for the maize flowering-related gene ZmRap2.7 and the tillering-related gene ZmTb1 demonstrated the feasibility of DeepCBA for exploration of regulatory elements that affect gene expression. Moreover, promoter editing and verification of two reported genes (ZmCLE7 and ZmVTE4) demonstrated the utility of DeepCBA for the precise design of gene expression and even for future intelligent breeding. DeepCBA is available at http://www.deepcba.com/ or http://124.220.197.196/.

A Functional Survey of the Regulatory Landscape of Estrogen Receptor-Positive Breast Cancer Evolution

Only a handful of somatic alterations have been linked to endocrine therapy resistance in hormone-dependent breast cancer, potentially explaining ∼40% of relapses. If other mechanisms underlie the evolution of hormone-dependent breast cancer under adjuvant therapy is currently unknown. In this work, we employ functional genomics to dissect the contribution of cis-regulatory elements (CRE) to cancer evolution by focusing on 12 megabases of noncoding DNA, including clonal enhancers, gene promoters, and boundaries of topologically associating domains. Parallel epigenetic perturbation (CRISPRi) in vitro reveals context-dependent roles for many of these CREs, with a specific impact on dormancy entrance and endocrine therapy resistance. Profiling of CRE somatic alterations in a unique, longitudinal cohort of patients treated with endocrine therapies identifies a limited set of noncoding changes potentially involved in therapy resistance. Overall, our data uncover how endocrine therapies trigger the emergence of transient features which could ultimately be exploited to hinder the adaptive process. Significance: This study shows that cells adapting to endocrine therapies undergo changes in the usage or regulatory regions. Dormant cells are less vulnerable to regulatory perturbation but gain transient dependencies which can be exploited to decrease the formation of dormant persisters.

Variant Impact Predictor database (VIPdb), version 2: trends from three decades of genetic variant impact predictors

BACKGROUND

Variant interpretation is essential for identifying patients' disease-causing genetic variants amongst the millions detected in their genomes. Hundreds of Variant Impact Predictors (VIPs), also known as Variant Effect Predictors (VEPs), have been developed for this purpose, with a variety of methodologies and goals. To facilitate the exploration of available VIP options, we have created the Variant Impact Predictor database (VIPdb).

RESULTS

The Variant Impact Predictor database (VIPdb) version 2 presents a collection of VIPs developed over the past three decades, summarizing their characteristics, ClinGen calibrated scores, CAGI assessment results, publication details, access information, and citation patterns. We previously summarized 217 VIPs and their features in VIPdb in 2019. Building upon this foundation, we identified and categorized an additional 190 VIPs, resulting in a total of 407 VIPs in VIPdb version 2. The majority of the VIPs have the capacity to predict the impacts of single nucleotide variants and nonsynonymous variants. More VIPs tailored to predict the impacts of insertions and deletions have been developed since the 2010s. In contrast, relatively few VIPs are dedicated to the prediction of splicing, structural, synonymous, and regulatory variants. The increasing rate of citations to VIPs reflects the ongoing growth in their use, and the evolving trends in citations reveal development in the field and individual methods.

CONCLUSIONS

VIPdb version 2 summarizes 407 VIPs and their features, potentially facilitating VIP exploration for various variant interpretation applications. VIPdb is available at  https://genomeinterpretation.org/vipdb.

Single-cell analysis of chromatin and expression reveals age- and sex-associated alterations in the human heart

Sex differences and age-related changes in the human heart at the tissue, cell, and molecular level have been well-documented and many may be relevant for cardiovascular disease. However, how molecular programs within individual cell types vary across individuals by age and sex remains poorly characterized. To better understand this variation, we performed single-nucleus combinatorial indexing (sci) ATAC- and RNA-Seq in human heart samples from nine donors. We identify hundreds of differentially expressed genes by age and sex and find epigenetic signatures of variation in ATAC-Seq data in this discovery cohort. We then scale up our single-cell RNA-Seq analysis by combining our data with five recently published single nucleus RNA-Seq datasets of healthy adult hearts. We find variation such as metabolic alterations by sex and immune changes by age in differential expression tests, as well as alterations in abundance of cardiomyocytes by sex and neurons with age. In addition, we compare our adult-derived ATAC-Seq profiles to analogous fetal cell types to identify putative developmental-stage-specific regulatory factors. Finally, we train predictive models of cell-type-specific RNA expression levels utilizing ATAC-Seq profiles to link distal regulatory sequences to promoters, quantifying the predictive value of a simple TF-to-expression regulatory grammar and identifying cell-type-specific TFs. Our analysis represents the largest single-cell analysis of cardiac variation by age and sex to date and provides a resource for further study of healthy cardiac variation and transcriptional regulation at single-cell resolution.

Transcription Attenuation in Synthetic Promoters in Nonoverlapping Tandem Formation

Closely spaced promoters are ubiquitous in prokaryotic and eukaryotic genomes. How their structure and dynamics relate remains unclear, particularly for tandem formations. To study their transcriptional interference, we engineered two pairs and one trio of synthetic promoters in nonoverlapping, tandem formation, in single-copy plasmids transformed into Escherichia coli cells. From in vivo measurements, we found that these promoters in tandem formation can have attenuated transcription rates. The attenuation strength can be widely fine-tuned by the promoters' positioning, natural regulatory mechanisms, and other factors, including the antibiotic rifampicin, which is known to hamper RNAP promoter escape. From this, and supported by in silico models, we concluded that the attenuation in these constructs emerges from premature terminations generated by collisions between RNAPs elongating from upstream promoters and RNAPs occupying downstream promoters. Moreover, we found that these collisions can cause one or both RNAPs to falloff. Finally, the broad spectrum of possible, externally regulated, attenuation strengths observed in our synthetic tandem promoters suggests that they could become useful as externally controllable regulators of future synthetic circuits.

Computational strategies for cross-species knowledge transfer and translational biomedicine

Research organisms provide invaluable insights into human biology and diseases, serving as essential tools for functional experiments, disease modeling, and drug testing. However, evolutionary divergence between humans and research organisms hinders effective knowledge transfer across species. Here, we review state-of-the-art methods for computationally transferring knowledge across species, primarily focusing on methods that utilize transcriptome data and/or molecular networks. We introduce the term "agnology" to describe the functional equivalence of molecular components regardless of evolutionary origin, as this concept is becoming pervasive in integrative data-driven models where the role of evolutionary origin can become unclear. Our review addresses four key areas of information and knowledge transfer across species: (1) transferring disease and gene annotation knowledge, (2) identifying agnologous molecular components, (3) inferring equivalent perturbed genes or gene sets, and (4) identifying agnologous cell types. We conclude with an outlook on future directions and several key challenges that remain in cross-species knowledge transfer.

Human genetics and epigenetics of alcohol use disorder

Alcohol use disorder (AUD) is a prominent contributor to global morbidity and mortality. Its complex etiology involves genetics, epigenetics, and environmental factors. We review progress in understanding the genetics and epigenetics of AUD, summarizing the key findings. Advancements in technology over the decades have elevated research from early candidate gene studies to present-day genome-wide scans, unveiling numerous genetic and epigenetic risk factors for AUD. The latest GWAS on more than one million participants identified more than 100 genetic variants, and the largest epigenome-wide association studies (EWAS) in blood and brain samples have revealed tissue-specific epigenetic changes. Downstream analyses revealed enriched pathways, genetic correlations with other traits, transcriptome-wide association in brain tissues, and drug-gene interactions for AUD. We also discuss limitations and future directions, including increasing the power of GWAS and EWAS studies as well as expanding the diversity of populations included in these analyses. Larger samples, novel technologies, and analytic approaches are essential; these include whole-genome sequencing, multiomics, single-cell sequencing, spatial transcriptomics, deep-learning prediction of variant function, and integrated methods for disease risk prediction.

Rare variation in non-coding regions with evolutionary signatures contributes to autism spectrum disorder risk

Little is known about the role of non-coding regions in the etiology of autism spectrum disorder (ASD). We examined three classes of non-coding regions: human accelerated regions (HARs), which show signatures of positive selection in humans; experimentally validated neural VISTA enhancers (VEs); and conserved regions predicted to act as neural enhancers (CNEs). Targeted and whole-genome analysis of >16,600 samples and >4,900 ASD probands revealed that likely recessive, rare, inherited variants in HARs, VEs, and CNEs substantially contribute to ASD risk in probands whose parents share ancestry, which enriches for recessive contributions, but modestly contribute, if at all, in simplex family structures. We identified multiple patient variants in HARs near IL1RAPL1 and in VEs near OTX1 and SIM1 and showed that they change enhancer activity. Our results implicate both human-evolved and evolutionarily conserved non-coding regions in ASD risk and suggest potential mechanisms of how regulatory changes can modulate social behavior.

Predicting the translation efficiency of messenger RNA in mammalian cells

The degree to which translational control is specified by mRNA sequence is poorly understood in mammalian cells. Here, we constructed and leveraged a compendium of 3,819 ribosomal profiling datasets, distilling them into a transcriptome-wide atlas of translation efficiency (TE) measurements encompassing >140 human and mouse cell types. We subsequently developed RiboNN, a multitask deep convolutional neural network, and classic machine learning models to predict TEs in hundreds of cell types from sequence-encoded mRNA features, achieving state-of-the-art performance (r=0.79 in human and r=0.78 in mouse for mean TE across cell types). While the majority of earlier models solely considered 5' UTR sequence, RiboNN integrates contributions from the full-length mRNA sequence, learning that the 5' UTR, CDS, and 3' UTR respectively possess ~67%, 31%, and 2% per-nucleotide information density in the specification of mammalian TEs. Interpretation of RiboNN revealed that the spatial positioning of low-level di- and tri-nucleotide features (i.e., including codons) largely explain model performance, capturing mechanistic principles such as how ribosomal processivity and tRNA abundance control translational output. RiboNN is predictive of the translational behavior of base-modified therapeutic RNA, and can explain evolutionary selection pressures in human 5' UTRs. Finally, it detects a common language governing mRNA regulatory control and highlights the interconnectedness of mRNA translation, stability, and localization in mammalian organisms.

Identification, Design, and Application of Noncoding Cis-Regulatory Elements

Cis-regulatory elements (CREs) play a pivotal role in orchestrating interactions with trans-regulatory factors such as transcription factors, RNA-binding proteins, and noncoding RNAs. These interactions are fundamental to the molecular architecture underpinning complex and diverse biological functions in living organisms, facilitating a myriad of sophisticated and dynamic processes. The rapid advancement in the identification and characterization of these regulatory elements has been marked by initiatives such as the Encyclopedia of DNA Elements (ENCODE) project, which represents a significant milestone in the field. Concurrently, the development of CRE detection technologies, exemplified by massively parallel reporter assays, has progressed at an impressive pace, providing powerful tools for CRE discovery. The exponential growth of multimodal functional genomic data has necessitated the application of advanced analytical methods. Deep learning algorithms, particularly large language models, have emerged as invaluable tools for deconstructing the intricate nucleotide sequences governing CRE function. These advancements facilitate precise predictions of CRE activity and enable the de novo design of CREs. A deeper understanding of CRE operational dynamics is crucial for harnessing their versatile regulatory properties. Such insights are instrumental in refining gene therapy techniques, enhancing the efficacy of selective breeding programs, pushing the boundaries of genetic innovation, and opening new possibilities in microbial synthetic biology.

EPInformer: a scalable deep learning framework for gene expression prediction by integrating promoter-enhancer sequences with multimodal epigenomic data

Transcriptional regulation, critical for cellular differentiation and adaptation to environmental changes, involves coordinated interactions among DNA sequences, regulatory proteins, and chromatin architecture. Despite extensive data from consortia like ENCODE, understanding the dynamics of cis-regulatory elements (CREs) in gene expression remains challenging. Deep learning is a powerful tool for learning gene expression and epigenomic signals from DNA sequences, exhibiting superior performance compared to conventional machine learning approaches. However, even the most advanced deep learning-based methods may fall short in capturing the regulatory effects of distal elements such as enhancers, limiting their predictive accuracy. In addition, these methods may require significant resources to train or to adapt to newly generated data. To address these challenges, we present EPInformer, a scalable deep-learning framework for predicting gene expression by integrating promoter-enhancer interactions with their sequences, epigenomic signals, and chromatin contacts. Our model outperforms existing gene expression prediction models in rigorous cross-chromosome validation, accurately recapitulates enhancer-gene interactions validated by CRISPR perturbation experiments, and identifies crucial transcription factor motifs within regulatory sequences. EPInformer is available as open-source software at https://github.com/pinellolab/EPInformer.

Deep Learning Sequence Models for Transcriptional Regulation

Deciphering the regulatory code of gene expression and interpreting the transcriptional effects of genome variation are critical challenges in human genetics. Modern experimental technologies have resulted in an abundance of data, enabling the development of sequence-based deep learning models that link patterns embedded in DNA to the biochemical and regulatory properties contributing to transcriptional regulation, including modeling epigenetic marks, 3D genome organization, and gene expression, with tissue and cell-type specificity. Such methods can predict the functional consequences of any noncoding variant in the human genome, even rare or never-before-observed variants, and systematically characterize their consequences beyond what is tractable from experiments or quantitative genetics studies alone. Recently, the development and application of interpretability approaches have led to the identification of key sequence patterns contributing to the predicted tasks, providing insights into the underlying biological mechanisms learned and revealing opportunities for improvement in future models.

Current genomic deep learning models display decreased performance in cell type-specific accessible regions

BACKGROUND

A number of deep learning models have been developed to predict epigenetic features such as chromatin accessibility from DNA sequence. Model evaluations commonly report performance genome-wide; however, cis regulatory elements (CREs), which play critical roles in gene regulation, make up only a small fraction of the genome. Furthermore, cell type-specific CREs contain a large proportion of complex disease heritability.

RESULTS

We evaluate genomic deep learning models in chromatin accessibility regions with varying degrees of cell type specificity. We assess two modeling directions in the field: general purpose models trained across thousands of outputs (cell types and epigenetic marks) and models tailored to specific tissues and tasks. We find that the accuracy of genomic deep learning models, including two state-of-the-art general purpose models-Enformer and Sei-varies across the genome and is reduced in cell type-specific accessible regions. Using accessibility models trained on cell types from specific tissues, we find that increasing model capacity to learn cell type-specific regulatory syntax-through single-task learning or high capacity multi-task models-can improve performance in cell type-specific accessible regions. We also observe that improving reference sequence predictions does not consistently improve variant effect predictions, indicating that novel strategies are needed to improve performance on variants.

CONCLUSIONS

Our results provide a new perspective on the performance of genomic deep learning models, showing that performance varies across the genome and is particularly reduced in cell type-specific accessible regions. We also identify strategies to maximize performance in cell type-specific accessible regions.

Deep learning approaches for non-coding genetic variant effect prediction: current progress and future prospects

Recent advancements in high-throughput sequencing technologies have significantly enhanced our ability to unravel the intricacies of gene regulatory processes. A critical challenge in this endeavor is the identification of variant effects, a key factor in comprehending the mechanisms underlying gene regulation. Non-coding variants, constituting over 90% of all variants, have garnered increasing attention in recent years. The exploration of gene variant impacts and regulatory mechanisms has spurred the development of various deep learning approaches, providing new insights into the global regulatory landscape through the analysis of extensive genetic data. Here, we provide a comprehensive overview of the development of the non-coding variants models based on bulk and single-cell sequencing data and their model-based interpretation and downstream tasks. This review delineates the popular sequencing technologies for epigenetic profiling and deep learning approaches for discerning the effects of non-coding variants. Additionally, we summarize the limitations of current approaches in variant effect prediction research and outline opportunities for improvement. We anticipate that our study will offer a practical and useful guide for the bioinformatic community to further advance the unraveling of genetic variant effects.

GENet: A Graph-Based Model Leveraging Histone Marks and Transcription Factors for Enhanced Gene Expression Prediction

Understanding the regulatory mechanisms of gene expression is a crucial objective in genomics. Although the DNA sequence near the transcription start site (TSS) offers valuable insights, recent methods suggest that analyzing only the surrounding DNA may not suffice to accurately predict gene expression levels. We developed GENet (Gene Expression Network from Histone and Transcription Factor Integration), a novel approach that integrates essential regulatory signals from transcription factors and histone modifications into a graph-based model. GENet extends beyond simple DNA sequence analysis by incorporating additional layers of genetic control, which are vital for determining gene expression. Our method markedly enhances the prediction of mRNA levels compared to previous models that depend solely on DNA sequence data. The results underscore the significance of including comprehensive regulatory information in gene expression studies. GENet emerges as a promising tool for researchers, with potential applications extending from fundamental biological research to the development of medical therapies.

DeepFace: Deep-learning-based framework to contextualize orofacial-cleft-related variants during human embryonic craniofacial development

Orofacial clefts (OFCs) are among the most common human congenital birth defects. Previous multiethnic studies have identified dozens of associated loci for both cleft lip with or without cleft palate (CL/P) and cleft palate alone (CP). Although several nearby genes have been highlighted, the "casual" variants are largely unknown. Here, we developed DeepFace, a convolutional neural network model, to assess the functional impact of variants by SNP activity difference (SAD) scores. The DeepFace model is trained with 204 epigenomic assays from crucial human embryonic craniofacial developmental stages of post-conception week (pcw) 4 to pcw 10. The Pearson correlation coefficient between the predicted and actual values for 12 epigenetic features achieved a median range of 0.50-0.83. Specifically, our model revealed that SNPs significantly associated with OFCs tended to exhibit higher SAD scores across various variant categories compared to less related groups, indicating a context-specific impact of OFC-related SNPs. Notably, we identified six SNPs with a significant linear relationship to SAD scores throughout developmental progression, suggesting that these SNPs could play a temporal regulatory role. Furthermore, our cell-type specificity analysis pinpointed the trophoblast cell as having the highest enrichment of risk signals associated with OFCs. Overall, DeepFace can harness distal regulatory signals from extensive epigenomic assays, offering new perspectives for prioritizing OFC variants using contextualized functional genomic features. We expect DeepFace to be instrumental in accessing and predicting the regulatory roles of variants associated with OFCs, and the model can be extended to study other complex diseases or traits.

Current genomic deep learning models display decreased performance in cell type specific accessible regions

Background

A number of deep learning models have been developed to predict epigenetic features such as chromatin accessibility from DNA sequence. Model evaluations commonly report performance genome-wide; however, cis regulatory elements (CREs), which play critical roles in gene regulation, make up only a small fraction of the genome. Furthermore, cell type specific CREs contain a large proportion of complex disease heritability.

Results

We evaluate genomic deep learning models in chromatin accessibility regions with varying degrees of cell type specificity. We assess two modeling directions in the field: general purpose models trained across thousands of outputs (cell types and epigenetic marks), and models tailored to specific tissues and tasks. We find that the accuracy of genomic deep learning models, including two state-of-the-art general purpose models - Enformer and Sei - varies across the genome and is reduced in cell type specific accessible regions. Using accessibility models trained on cell types from specific tissues, we find that increasing model capacity to learn cell type specific regulatory syntax - through single-task learning or high capacity multi-task models - can improve performance in cell type specific accessible regions. We also observe that improving reference sequence predictions does not consistently improve variant effect predictions, indicating that novel strategies are needed to improve performance on variants.

Conclusions

Our results provide a new perspective on the performance of genomic deep learning models, showing that performance varies across the genome and is particularly reduced in cell type specific accessible regions. We also identify strategies to maximize performance in cell type specific accessible regions.

A foundational large language model for edible plant genomes

Significant progress has been made in the field of plant genomics, as demonstrated by the increased use of high-throughput methodologies that enable the characterization of multiple genome-wide molecular phenotypes. These findings have provided valuable insights into plant traits and their underlying genetic mechanisms, particularly in model plant species. Nonetheless, effectively leveraging them to make accurate predictions represents a critical step in crop genomic improvement. We present AgroNT, a foundational large language model trained on genomes from 48 plant species with a predominant focus on crop species. We show that AgroNT can obtain state-of-the-art predictions for regulatory annotations, promoter/terminator strength, tissue-specific gene expression, and prioritize functional variants. We conduct a large-scale in silico saturation mutagenesis analysis on cassava to evaluate the regulatory impact of over 10 million mutations and provide their predicted effects as a resource for variant characterization. Finally, we propose the use of the diverse datasets compiled here as the Plants Genomic Benchmark (PGB), providing a comprehensive benchmark for deep learning-based methods in plant genomic research. The pre-trained AgroNT model is publicly available on HuggingFace at https://huggingface.co/InstaDeepAI/agro-nucleotide-transformer-1b  for future research purposes.

The contribution of silencer variants to human diseases

BACKGROUND

Although disease-causal genetic variants have been found within silencer sequences, we still lack a comprehensive analysis of the association of silencers with diseases. Here, we profiled GWAS variants in 2.8 million candidate silencers across 97 human samples derived from a diverse panel of tissues and developmental time points, using deep learning models.

RESULTS

We show that candidate silencers exhibit strong enrichment in disease-associated variants, and several diseases display a much stronger association with silencer variants than enhancer variants. Close to 52% of candidate silencers cluster, forming silencer-rich loci, and, in the loci of Parkinson's-disease-hallmark genes TRIM31 and MAL, the associated SNPs densely populate clustered candidate silencers rather than enhancers displaying an overall twofold enrichment in silencers versus enhancers. The disruption of apoptosis in neuronal cells is associated with both schizophrenia and bipolar disorder and can largely be attributed to variants within candidate silencers. Our model permits a mechanistic explanation of causative SNP effects by identifying altered binding of tissue-specific repressors and activators, validated with a 70% of directional concordance using SNP-SELEX. Narrowing the focus of the analysis to individual silencer variants, experimental data confirms the role of the rs62055708 SNP in Parkinson's disease, rs2535629 in schizophrenia, and rs6207121 in type 1 diabetes.

CONCLUSIONS

In summary, our results indicate that advances in deep learning models for the discovery of disease-causal variants within candidate silencers effectively "double" the number of functionally characterized GWAS variants. This provides a basis for explaining mechanisms of action and designing novel diagnostics and therapeutics.

An atlas of transcribed enhancers across helper T cell diversity for decoding human diseases

Transcribed enhancer maps can reveal nuclear interactions underpinning each cell type and connect specific cell types to diseases. Using a 5' single-cell RNA sequencing approach, we defined transcription start sites of enhancer RNAs and other classes of coding and noncoding RNAs in human CD4+ T cells, revealing cellular heterogeneity and differentiation trajectories. Integration of these datasets with single-cell chromatin profiles showed that active enhancers with bidirectional RNA transcription are highly cell type-specific and that disease heritability is strongly enriched in these enhancers. The resulting cell type-resolved multimodal atlas of bidirectionally transcribed enhancers, which we linked with promoters using fine-scale chromatin contact maps, enabled us to systematically interpret genetic variants associated with a range of immune-mediated diseases.

GET: a foundation model of transcription across human cell types

Transcriptional regulation, involving the complex interplay between regulatory sequences and proteins, directs all biological processes. Computational models of transcription lack generalizability to accurately extrapolate in unseen cell types and conditions. Here, we introduce GET, an interpretable foundation model designed to uncover regulatory grammars across 213 human fetal and adult cell types. Relying exclusively on chromatin accessibility data and sequence information, GET achieves experimental-level accuracy in predicting gene expression even in previously unseen cell types. GET showcases remarkable adaptability across new sequencing platforms and assays, enabling regulatory inference across a broad range of cell types and conditions, and uncovering universal and cell type specific transcription factor interaction networks. We evaluated its performance on prediction of regulatory activity, inference of regulatory elements and regulators, and identification of physical interactions between transcription factors. Specifically, we show GET outperforms current models in predicting lentivirus-based massive parallel reporter assay readout with reduced input data. In fetal erythroblasts, we identify distal (>1Mbp) regulatory regions that were missed by previous models. In B cells, we identified a lymphocyte-specific transcription factor-transcription factor interaction that explains the functional significance of a leukemia-risk predisposing germline mutation. In sum, we provide a generalizable and accurate model for transcription together with catalogs of gene regulation and transcription factor interactions, all with cell type specificity.

Analysis of RNA translation with a deep learning architecture provides new insight into translation control

Accurate annotation of coding regions in RNAs is essential for understanding gene translation. We developed a deep neural network to directly predict and analyze translation initiation and termination sites from RNA sequences. Trained with human transcripts, our model learned hidden rules of translation control and achieved a near perfect prediction of canonical translation sites across entire human transcriptome. Surprisingly, this model revealed a new role of codon usage in regulating translation termination, which was experimentally validated. We also identified thousands of new open reading frames in mRNAs or lncRNAs, some of which were confirmed experimentally. The model trained with human mRNAs achieved high prediction accuracy of canonical translation sites in all eukaryotes and good prediction in polycistronic transcripts from prokaryotes or RNA viruses, suggesting a high degree of conservation in translation control. Collectively, we present a general and efficient deep learning model for RNA translation, generating new insights into the complexity of translation regulation.

Advances in computational and experimental approaches for deciphering transcriptional regulatory networks: Understanding the roles of cis-regulatory elements is essential, and recent research utilizing MPRAs, STARR-seq, CRISPR-Cas9, and machine learning has yielded valuable insights

Understanding the influence of cis-regulatory elements on gene regulation poses numerous challenges given complexities stemming from variations in transcription factor (TF) binding, chromatin accessibility, structural constraints, and cell-type differences. This review discusses the role of gene regulatory networks in enhancing understanding of transcriptional regulation and covers construction methods ranging from expression-based approaches to supervised machine learning. Additionally, key experimental methods, including MPRAs and CRISPR-Cas9-based screening, which have significantly contributed to understanding TF binding preferences and cis-regulatory element functions, are explored. Lastly, the potential of machine learning and artificial intelligence to unravel cis-regulatory logic is analyzed. These computational advances have far-reaching implications for precision medicine, therapeutic target discovery, and the study of genetic variations in health and disease.

Variant Impact Predictor database (VIPdb), version 2: Trends from 25 years of genetic variant impact predictors

Background

Variant interpretation is essential for identifying patients' disease-causing genetic variants amongst the millions detected in their genomes. Hundreds of Variant Impact Predictors (VIPs), also known as Variant Effect Predictors (VEPs), have been developed for this purpose, with a variety of methodologies and goals. To facilitate the exploration of available VIP options, we have created the Variant Impact Predictor database (VIPdb).

Results

The Variant Impact Predictor database (VIPdb) version 2 presents a collection of VIPs developed over the past 25 years, summarizing their characteristics, ClinGen calibrated scores, CAGI assessment results, publication details, access information, and citation patterns. We previously summarized 217 VIPs and their features in VIPdb in 2019. Building upon this foundation, we identified and categorized an additional 186 VIPs, resulting in a total of 403 VIPs in VIPdb version 2. The majority of the VIPs have the capacity to predict the impacts of single nucleotide variants and nonsynonymous variants. More VIPs tailored to predict the impacts of insertions and deletions have been developed since the 2010s. In contrast, relatively few VIPs are dedicated to the prediction of splicing, structural, synonymous, and regulatory variants. The increasing rate of citations to VIPs reflects the ongoing growth in their use, and the evolving trends in citations reveal development in the field and individual methods.

Conclusions

VIPdb version 2 summarizes 403 VIPs and their features, potentially facilitating VIP exploration for various variant interpretation applications.

Availability

VIPdb version 2 is available at https://genomeinterpretation.org/vipdb.

Modeling 0.6 million genes for the rational design of functional cis-regulatory variants and de novo design of cis-regulatory sequences

Rational design of plant cis-regulatory DNA sequences without expert intervention or prior domain knowledge is still a daunting task. Here, we developed PhytoExpr, a deep learning framework capable of predicting both mRNA abundance and plant species using the proximal regulatory sequence as the sole input. PhytoExpr was trained over 17 species representative of major clades of the plant kingdom to enhance its generalizability. Via input perturbation, quantitative functional annotation of the input sequence was achieved at single-nucleotide resolution, revealing an abundance of predicted high-impact nucleotides in conserved noncoding sequences and transcription factor binding sites. Evaluation of maize HapMap3 single-nucleotide polymorphisms (SNPs) by PhytoExpr demonstrates an enrichment of predicted high-impact SNPs in cis-eQTL. Additionally, we provided two algorithms that harnessed the power of PhytoExpr in designing functional cis-regulatory variants, and de novo creation of species-specific cis-regulatory sequences through in silico evolution of random DNA sequences. Our model represents a general and robust approach for functional variant discovery in population genetics and rational design of regulatory sequences for genome editing and synthetic biology.

SuPreMo: a computational tool for streamlining in silico perturbation using sequence-based predictive models

SUMMARY

The increasing development of sequence-based machine learning models has raised the demand for manipulating sequences for this application. However, existing approaches to edit and evaluate genome sequences using models have limitations, such as incompatibility with structural variants, challenges in identifying responsible sequence perturbations, and the need for vcf file inputs and phased data. To address these bottlenecks, we present Sequence Mutator for Predictive Models (SuPreMo), a scalable and comprehensive tool for performing and supporting in silico mutagenesis experiments. We then demonstrate how pairs of reference and perturbed sequences can be used with machine learning models to prioritize pathogenic variants or discover new functional sequences.

AVAILABILITY AND IMPLEMENTATION

SuPreMo was written in Python, and can be run using only one line of code to generate both sequences and 3D genome disruption scores. The codebase, instructions for installation and use, and tutorials are on the GitHub page: https://github.com/ketringjoni/SuPreMo.

Big data and deep learning for RNA biology

The exponential growth of big data in RNA biology (RB) has led to the development of deep learning (DL) models that have driven crucial discoveries. As constantly evidenced by DL studies in other fields, the successful implementation of DL in RB depends heavily on the effective utilization of large-scale datasets from public databases. In achieving this goal, data encoding methods, learning algorithms, and techniques that align well with biological domain knowledge have played pivotal roles. In this review, we provide guiding principles for applying these DL concepts to various problems in RB by demonstrating successful examples and associated methodologies. We also discuss the remaining challenges in developing DL models for RB and suggest strategies to overcome these challenges. Overall, this review aims to illuminate the compelling potential of DL for RB and ways to apply this powerful technology to investigate the intriguing biology of RNA more effectively.

CAPTURE: Comprehensive anti-cancer peptide predictor with a unique amino acid sequence encoder

Anticancer peptides (ACPs) key properties including bioactivity, high efficacy, low toxicity, and lack of drug resistance make them ideal candidates for cancer therapies. To deeply explore the potential of ACPs and accelerate development of cancer therapies, although 53 Artificial Intelligence supported computational predictors have been developed for ACPs and non ACPs classification but only one predictor has been developed for ACPs functional types annotations. Moreover, these predictors extract amino acids distribution patterns to transform peptides sequences into statistical vectors that are further fed to classifiers for discriminating peptides sequences and annotating peptides functional classes. Overall, these predictors remain fail in extracting diverse types of amino acids distribution patterns from peptide sequences. The paper in hand presents a unique CARE encoder that transforms peptides sequences into statistical vectors by extracting 4 different types of distribution patterns including correlation, distribution, composition, and transition. Across public benchmark dataset, proposed encoder potential is explored under two different evaluation settings namely; intrinsic and extrinsic. Extrinsic evaluation indicates that 12 different machine learning classifiers achieve superior performance with the proposed encoder as compared to 55 existing encoders. Furthermore, an intrinsic evaluation reveals that, unlike existing encoders, the proposed encoder generates more discriminative clusters for ACPs and non-ACPs classes. Across 8 public benchmark ACPs and non-ACPs classification datasets, proposed encoder and Adaboost classifier based CAPTURE predictor outperforms existing predictors with an average accuracy, recall and MCC score of 1%, 4%, and 2% respectively. In generalizeability evaluation case study, across 7 benchmark anti-microbial peptides classification datasets, CAPTURE surpasses existing predictors by an average AU-ROC of 2%. CAPTURE predictive pipeline along with label powerset method outperforms state-of-the-art ACPs functional types predictor by 5%, 5%, 5%, 6%, and 3% in terms of average accuracy, subset accuracy, precision, recall, and F1 respectively. CAPTURE web application is available at https://sds_genetic_analysis.opendfki.de/CAPTURE.

Decoding polygenic diseases: advances in noncoding variant prioritization and validation

Genome-wide association studies (GWASs) provide a key foundation for elucidating the genetic underpinnings of common polygenic diseases. However, these studies have limitations in their ability to assign causality to particular genetic variants, especially those residing in the noncoding genome. Over the past decade, technological and methodological advances in both analytical and empirical prioritization of noncoding variants have enabled the identification of causative variants by leveraging orthogonal functional evidence at increasing scale. In this review, we present an overview of these approaches and describe how this workflow provides the groundwork necessary to move beyond associations toward genetically informed studies on the molecular and cellular mechanisms of polygenic disease.

Validated Negative Regions (VNRs) in the VISTA Database might be Truncated Forms of Bona Fide Enhancers

The VISTA enhancer database is a valuable resource for evaluating predicted enhancers in humans and mice. In addition to thousands of validated positive regions (VPRs) in the human and mouse genomes, the database also contains similar numbers of validated negative regions (VNRs). It is previously shown that the VPRs are on average half as long as predicted overlapping enhancers that are highly conserved and hypothesize that the VPRs may be truncated forms of long bona fide enhancers. Here, it is shown that like the VPRs, the VNRs also are under strong evolutionary constraints and overlap predicted enhancers in the genomes. The VNRs are also on average half as long as predicted overlapping enhancers that are highly conserved. Moreover, the VNRs and the VPRs display similar cell/tissue-specific modification patterns of key epigenetic marks of active enhancers. Furthermore, the VNRs and the VPRs show similar impact score spectra of in silico mutagenesis. These highly similar properties between the VPRs and the VNRs suggest that like the VPRs, the VNRs may also be truncated forms of long bona fide enhancers.

Interpretable prediction of mRNA abundance from promoter sequence using contextual regression models

While machine learning models have been successfully applied to predicting gene expression from promoter sequences, it remains a great challenge to derive intuitive interpretation of the model and reveal DNA motif grammar such as motif cooperation and distance constraint between motif sites. Previous interpretation approaches are often time-consuming or have difficulty to learn the combinatory rules. In this work, we designed interpretable neural network models to predict the mRNA expression levels from DNA sequences. By applying the Contextual Regression framework we developed, we extracted weighted features to cluster samples into different groups, which have different gene expression levels. We performed motif analysis in each cluster and found motifs with active or repressive regulation on gene expression. By comparing the co-occurrence locations of discovered motifs, we also uncovered multiple grammars of motif combination including communities of cooperative motifs and distance constraints between motif pairs. These results revealed new insights of the regulatory architecture of promoter sequences.

Dissecting the cis-regulatory syntax of transcription initiation with deep learning

Despite extensive characterization of mammalian Pol II transcription, the DNA sequence determinants of transcription initiation at a third of human promoters and most enhancers remain poorly understood. Hence, we trained and interpreted a neural network called ProCapNet that accurately models base-resolution initiation profiles from PRO-cap experiments using local DNA sequence. ProCapNet learns sequence motifs with distinct effects on initiation rates and TSS positioning and uncovers context-specific cryptic initiator elements intertwined within other TF motifs. ProCapNet annotates predictive motifs in nearly all actively transcribed regulatory elements across multiple cell-lines, revealing a shared cis-regulatory logic across promoters and enhancers mediated by a highly epistatic sequence syntax of cooperative and competitive motif interactions. ProCapNet models of RAMPAGE profiles measuring steady-state RNA abundance at TSSs distill initiation signals on par with models trained directly on PRO-cap profiles. ProCapNet learns a largely cell-type-agnostic cis-regulatory code of initiation complementing sequence drivers of cell-type-specific chromatin state critical for accurate prediction of cell-type-specific transcription initiation.

Deconvolution of polygenic risk score in single cells unravels cellular and molecular heterogeneity of complex human diseases

Polygenic risk scores (PRSs) are commonly used for predicting an individual's genetic risk of complex diseases. Yet, their implication for disease pathogenesis remains largely limited. Here, we introduce scPRS, a geometric deep learning model that constructs single-cell-resolved PRS leveraging reference single-cell chromatin accessibility profiling data to enhance biological discovery as well as disease prediction. Real-world applications across multiple complex diseases, including type 2 diabetes (T2D), hypertrophic cardiomyopathy (HCM), and Alzheimer's disease (AD), showcase the superior prediction power of scPRS compared to traditional PRS methods. Importantly, scPRS not only predicts disease risk but also uncovers disease-relevant cells, such as hormone-high alpha and beta cells for T2D, cardiomyocytes and pericytes for HCM, and astrocytes, microglia and oligodendrocyte progenitor cells for AD. Facilitated by a layered multi-omic analysis, scPRS further identifies cell-type-specific genetic underpinnings, linking disease-associated genetic variants to gene regulation within corresponding cell types. We substantiate the disease relevance of scPRS-prioritized HCM genes and demonstrate that the suppression of these genes in HCM cardiomyocytes is rescued by Mavacamten treatment. Additionally, we establish a novel microglia-specific regulatory relationship between the AD risk variant rs7922621 and its target genes ANXA11 and TSPAN14. We further illustrate the detrimental effects of suppressing these two genes on microglia phagocytosis. Our work provides a multi-tasking, interpretable framework for precise disease prediction and systematic investigation of the genetic, cellular, and molecular basis of complex diseases, laying the methodological foundation for single-cell genetics.

Functional dissection of complex and molecular trait variants at single nucleotide resolution

Identifying the causal variants and mechanisms that drive complex traits and diseases remains a core problem in human genetics. The majority of these variants have individually weak effects and lie in non-coding gene-regulatory elements where we lack a complete understanding of how single nucleotide alterations modulate transcriptional processes to affect human phenotypes. To address this, we measured the activity of 221,412 trait-associated variants that had been statistically fine-mapped using a Massively Parallel Reporter Assay (MPRA) in 5 diverse cell-types. We show that MPRA is able to discriminate between likely causal variants and controls, identifying 12,025 regulatory variants with high precision. Although the effects of these variants largely agree with orthogonal measures of function, only 69% can plausibly be explained by the disruption of a known transcription factor (TF) binding motif. We dissect the mechanisms of 136 variants using saturation mutagenesis and assign impacted TFs for 91% of variants without a clear canonical mechanism. Finally, we provide evidence that epistasis is prevalent for variants in close proximity and identify multiple functional variants on the same haplotype at a small, but important, subset of trait-associated loci. Overall, our study provides a systematic functional characterization of likely causal common variants underlying complex and molecular human traits, enabling new insights into the regulatory grammar underlying disease risk.

How DNA encodes the start of transcription

A deep-learning model reveals the rules that define transcription initiation.

Sequence basis of transcription initiation in the human genome

Transcription initiation is a process that is essential to ensuring the proper function of any gene, yet we still lack a unified understanding of sequence patterns and rules that explain most transcription start sites in the human genome. By predicting transcription initiation at base-pair resolution from sequences with a deep learning-inspired explainable model called Puffin, we show that a small set of simple rules can explain transcription initiation at most human promoters. We identify key sequence patterns that contribute to human promoter activity, each activating transcription with distinct position-specific effects. Furthermore, we explain the sequence basis of bidirectional transcription at promoters, identify the links between promoter sequence and gene expression variation across cell types, and explore the conservation of sequence determinants of transcription initiation across mammalian species.

Haplotype function score improves biological interpretation and cross-ancestry polygenic prediction of human complex traits

We propose a new framework for human genetic association studies: at each locus, a deep learning model (in this study, Sei) is used to calculate the functional genomic activity score for two haplotypes per individual. This score, defined as the Haplotype Function Score (HFS), replaces the original genotype in association studies. Applying the HFS framework to 14 complex traits in the UK Biobank, we identified 3619 independent HFS-trait associations with a significance of p < 5 × 10-8. Fine-mapping revealed 2699 causal associations, corresponding to a median increase of 63 causal findings per trait compared with single-nucleotide polymorphism (SNP)-based analysis. HFS-based enrichment analysis uncovered 727 pathway-trait associations and 153 tissue-trait associations with strong biological interpretability, including 'circadian pathway-chronotype' and 'arachidonic acid-intelligence'. Lastly, we applied least absolute shrinkage and selection operator (LASSO) regression to integrate HFS prediction score with SNP-based polygenic risk scores, which showed an improvement of 16.1-39.8% in cross-ancestry polygenic prediction. We concluded that HFS is a promising strategy for understanding the genetic basis of human complex traits.