A Genome-wide Framework for Mapping Gene Regulation via Cellular Genetic Screens.

Over one million candidate regulatory elements have been identified across the human genome, but nearly all are unvalidated and their target genes uncertain. Approaches based on human genetics are limited in scope to common variants and in resolution by linkage disequilibrium. We present a multiplex, expression quantitative trait locus (eQTL)-inspired framework for mapping enhancer-gene pairs by introducing random combinations of CRISPR/Cas9-mediated perturbations to each of many cells, followed by single-cell RNA sequencing (RNA-seq). Across two experiments, we used dCas9-KRAB to perturb 5,920 candidate enhancers with no strong a priori hypothesis as to their target gene(s), measuring effects by profiling 254,974 single-cell transcriptomes. We identified 664 (470 high-confidence) cis enhancer-gene pairs, which were enriched for specific transcription factors, non-housekeeping status, and genomic and 3D conformational proximity to their target genes. This framework will facilitate the large-scale mapping of enhancer-gene regulatory interactions, a critical yet largely uncharted component of the cis-regulatory landscape of the human genome.

Navigating the pitfalls of applying machine learning in genomics.

The scale of genetic, epigenomic, transcriptomic, cheminformatic and proteomic data available today, coupled with easy-to-use machine learning (ML) toolkits, has propelled the application of supervised learning in genomics research. However, the assumptions behind the statistical models and performance evaluations in ML software frequently are not met in biological systems. In this Review, we illustrate the impact of several common pitfalls encountered when applying supervised ML in genomics. We explore how the structure of genomics data can bias performance evaluations and predictions. To address the challenges associated with applying cutting-edge ML methods to genomics, we describe solutions and appropriate use cases where ML modelling shows great potential.

A High-Throughput Screen for Transcription Activation Domains Reveals Their Sequence Features and Permits Prediction by Deep Learning.

Acidic transcription activation domains (ADs) are encoded by a wide range of seemingly unrelated amino acid sequences, making it difficult to recognize features that promote their dynamic behavior, "fuzzy" interactions, and target specificity. We screened a large set of random 30-mer peptides for AD function in yeast and trained a deep neural network (ADpred) on the AD-positive and -negative sequences. ADpred identifies known acidic ADs within transcription factors and accurately predicts the consequences of mutations. Our work reveals that strong acidic ADs contain multiple clusters of hydrophobic residues near acidic side chains, explaining why ADs often have a biased amino acid composition. ADs likely use a binding mechanism similar to avidity where a minimum number of weak dynamic interactions are required between activator and target to generate biologically relevant affinity and in vivo function. This mechanism explains the basis for fuzzy binding observed between acidic ADs and targets.

GENCODE: reference annotation for the human and mouse genomes in 2023.

GENCODE produces high quality gene and transcript annotation for the human and mouse genomes. All GENCODE annotation is supported by experimental data and serves as a reference for genome biology and clinical genomics. The GENCODE consortium generates targeted experimental data, develops bioinformatic tools and carries out analyses that, along with externally produced data and methods, support the identification and annotation of transcript structures and the determination of their function. Here, we present an update on the annotation of human and mouse genes, including developments in the tools, data, analyses and major collaborations which underpin this progress. For example, we report the creation of a set of non-canonical ORFs identified in GENCODE transcripts, the LRGASP collaboration to assess the use of long transcriptomic data to build transcript models, the progress in collaborations with RefSeq and UniProt to increase convergence in the annotation of human and mouse protein-coding genes, the propagation of GENCODE across the human pan-genome and the development of new tools to support annotation of regulatory features by GENCODE. Our annotation is accessible via Ensembl, the UCSC Genome Browser and https://www.gencodegenes.org.

Leveraging epigenomes and three-dimensional genome organization for interpreting regulatory variation.

Understanding the impact of regulatory variants on complex phenotypes is a significant challenge because the genes and pathways that are targeted by such variants and the cell type context in which regulatory variants operate are typically unknown. Cell-type-specific long-range regulatory interactions that occur between a distal regulatory sequence and a gene offer a powerful framework for examining the impact of regulatory variants on complex phenotypes. However, high-resolution maps of such long-range interactions are available only for a handful of cell types. Furthermore, identifying specific gene subnetworks or pathways that are targeted by a set of variants is a significant challenge. We have developed L-HiC-Reg, a Random Forests regression method to predict high-resolution contact counts in new cell types, and a network-based framework to identify candidate cell-type-specific gene networks targeted by a set of variants from a genome-wide association study (GWAS). We applied our approach to predict interactions in 55 Roadmap Epigenomics Mapping Consortium cell types, which we used to interpret regulatory single nucleotide polymorphisms (SNPs) in the NHGRI-EBI GWAS catalogue. Using our approach, we performed an in-depth characterization of fifteen different phenotypes including schizophrenia, coronary artery disease (CAD) and Crohn's disease. We found differentially wired subnetworks consisting of known as well as novel gene targets of regulatory SNPs. Taken together, our compendium of interactions and the associated network-based analysis pipeline leverages long-range regulatory interactions to examine the context-specific impact of regulatory variation in complex phenotypes.

Accelerating in silico saturation mutagenesis using compressed sensing.

MOTIVATION: In silico saturation mutagenesis (ISM) is a popular approach in computational genomics for calculating feature attributions on biological sequences that proceeds by systematically perturbing each position in a sequence and recording the difference in model output. However, this method can be slow because systematically perturbing each position requires performing a number of forward passes proportional to the length of the sequence being examined. RESULTS: In this work, we propose a modification of ISM that leverages the principles of compressed sensing to require only a constant number of forward passes, regardless of sequence length, when applied to models that contain operations with a limited receptive field, such as convolutions. Our method, named Yuzu, can reduce the time that ISM spends in convolution operations by several orders of magnitude and, consequently, Yuzu can speed up ISM on several commonly used architectures in genomics by over an order of magnitude. Notably, we found that Yuzu provides speedups that increase with the complexity of the convolution operation and the length of the sequence being analyzed, suggesting that Yuzu provides large benefits in realistic settings. AVAILABILITY AND IMPLEMENTATION: We have made this tool available at https://github.com/kundajelab/yuzu.

fastISM: performant in silico saturation mutagenesis for convolutional neural networks.

MOTIVATION: Deep-learning models, such as convolutional neural networks, are able to accurately map biological sequences to associated functional readouts and properties by learning predictive de novo representations. In silico saturation mutagenesis (ISM) is a popular feature attribution technique for inferring contributions of all characters in an input sequence to the model's predicted output. The main drawback of ISM is its runtime, as it involves multiple forward propagations of all possible mutations of each character in the input sequence through the trained model to predict the effects on the output. RESULTS: We present fastISM, an algorithm that speeds up ISM by a factor of over 10× for commonly used convolutional neural network architectures. fastISM is based on the observations that the majority of computation in ISM is spent in convolutional layers, and a single mutation only disrupts a limited region of intermediate layers, rendering most computation redundant. fastISM reduces the gap between backpropagation-based feature attribution methods and ISM. It far surpasses the runtime of backpropagation-based methods on multi-output architectures, making it feasible to run ISM on a large number of sequences. AVAILABILITY AND IMPLEMENTATION: An easy-to-use Keras/TensorFlow 2 implementation of fastISM is available at https://github.com/kundajelab/fastISM. fastISM can be installed using pip install fastism. A hands-on tutorial can be found at https://colab.research.google.com/github/kundajelab/fastISM/blob/master/notebooks/colab/DeepSEA.ipynb. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Prioritizing transcriptomic and epigenomic experiments using an optimization strategy that leverages imputed data.

MOTIVATION: Successful science often involves not only performing experiments well, but also choosing well among many possible experiments. In a hypothesis generation setting, choosing an experiment well means choosing an experiment whose results are interesting or novel. In this work, we formalize this selection procedure in the context of genomics and epigenomics data generation. Specifically, we consider the task faced by a scientific consortium such as the National Institutes of Health ENCODE Consortium, whose goal is to characterize all of the functional elements in the human genome. Given a list of possible cell types or tissue types ('biosamples') and a list of possible high-throughput sequencing assays, where at least one experiment has been performed in each biosample and for each assay, we ask 'Which experiments should ENCODE perform next?' RESULTS: We demonstrate how to represent this task as a submodular optimization problem, where the goal is to choose a panel of experiments that maximize the facility location function. A key aspect of our approach is that we use imputed data, rather than experimental data, to directly answer the posed question. We find that, across several evaluations, our method chooses a panel of experiments that span a diversity of biochemical activity. Finally, we propose two modifications of the facility location function, including a novel submodular-supermodular function, that allow incorporation of domain knowledge or constraints into the optimization procedure. AVAILABILITY AND IMPLEMENTATION: Our method is available as a Python package at https://github.com/jmschrei/kiwano and can be installed using the command pip install kiwano. The source code used here and the similarity matrix can be found at http://doi.org/10.5281/zenodo.3708538. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Medical student anxiety and depression in the COVID-19 Era: Unique needs of underrepresented students.

Background: The COVID-19 pandemic has caused significant morbidity, mortality, and mental health consequences. Few studies have examined the mental toll of COVID-19 on United States (US) medical students, who experience greater rates of depression and anxiety than the general population. Students who identify as underrepresented in medicine (URM) may experience even greater mental health adversities than non-URM peers. This study examines COVID-19's impact on preclinical medical student anxiety and depression and unique challenges disproportionately affecting URM students during the initial phase of the pandemic. Methods: Medical students at four US institutions completed an anonymous survey including the Patient Health Questionnaire-9 (PHQ-9) and Generalized Anxiety Disorder-7 (GAD-7) questionnaires for depression and anxiety. Participants provided information on demographics, past mental health difficulties, and concerns during the pandemic. Chi-square and Mann-Whitney U tests were performed using SPSS. Results: During the initial phase of the pandemic, URMs were 3.71 times more likely to be in the at-risk category on GAD-7 than non-URM peers. Before COVID-19, there was no significant difference between self-reported feelings or diagnoses of anxiety between groups. During the COVID-19 pandemic, there were significant differences in feelings of increased anxiety between URM (Mdn = 76) and non-URM (Mdn = 49) students, U = 702.5, P < 0.001, feelings of increased sadness between URM (Mdn = 49) and non-URM (Mdn = 34) students, U = 1036.5, P = 0.042, concern for new financial difficulty between URM (Mdn = 50) and non-URM students (Mdn = 7), U = 950.5, P = 0.012, and concern about lack of mental health support from their academic institution between URM (Mdn = 18) and non-URM students (Mdn = 9), U = 1083, P = 0.036 (one-tailed). Discussion: Large-scale crises such as COVID-19 may exacerbate mental health disparities between URM and non-URM students. Medical schools should consider increasing financial and mental health support for URM students in response to these significant adverse events.

Massively parallel profiling and predictive modeling of the outcomes of CRISPR/Cas9-mediated double-strand break repair.

Non-homologous end-joining (NHEJ) plays an important role in double-strand break (DSB) repair of DNA. Recent studies have shown that the error patterns of NHEJ are strongly biased by sequence context, but these studies were based on relatively few templates. To investigate this more thoroughly, we systematically profiled ∼1.16 million independent mutational events resulting from CRISPR/Cas9-mediated cleavage and NHEJ-mediated DSB repair of 6872 synthetic target sequences, introduced into a human cell line via lentiviral infection. We find that: (i) insertions are dominated by 1 bp events templated by sequence immediately upstream of the cleavage site, (ii) deletions are predominantly associated with microhomology and (iii) targets exhibit variable but reproducible diversity with respect to the number and relative frequency of the mutational outcomes to which they give rise. From these data, we trained a model that uses local sequence context to predict the distribution of mutational outcomes. Exploiting the bias of NHEJ outcomes towards microhomology mediated events, we demonstrate the programming of deletion patterns by introducing microhomology to specific locations in the vicinity of the DSB site. We anticipate that our results will inform investigations of DSB repair mechanisms as well as the design of CRISPR/Cas9 experiments for diverse applications including genome-wide screens, gene therapy, lineage tracing and molecular recording.

Maladaptive Perfectionism, Impostor Phenomenon, and Suicidal Ideation Among Medical Students.

OBJECTIVE: Suicide is a leading cause of death for young adults, and medical students experience elevated rates of suicide and suicidal ideation. The present study uses mediation analysis to explore relationships between suicidal ideation and two dysfunctional mindsets common among medical students: maladaptive perfectionism, high standards accompanied by excessive self-criticism, and impostor phenomenon, pervasive feelings of inadequacy despite evidence of competence and success. METHODS: Two hundred and twenty-six medical students at a single institution completed an online survey which assessed maladaptive perfectionism, impostor phenomenon, and suicidal ideation. After calculating measures of association between all study variables, linear regression was conducted to establish the relationship between maladaptive perfectionism and suicidal ideation. To evaluate whether impostor phenomenon mediated the relationship between maladaptive perfectionism and suicidal ideation as hypothesized, a series of regression models were constructed and the regression coefficients were examined. The statistical significance of the indirect effect, representing the mediated relationship, was tested using bootstrapping. RESULTS: Significant positive associations between maladaptive perfectionism, impostor phenomenon, and suicidal ideation were observed. Impostor phenomenon score was found to mediate the relationship between maladaptive perfectionism and suicidal ideation. CONCLUSIONS: Medical students who exhibit maladaptive perfectionism are at increased risk for feelings of impostor phenomenon, which translates into increased risk for suicide. These results suggest that an intervention targeted at reducing feelings of impostor phenomenon among maladaptive perfectionists may be effective in reducing their higher risk for suicide. However, interventions promoting individual resilience are not sufficient; systemic change is needed to address medicine's "culture of perfection."

Analysis of nanopore data using hidden Markov models.

MOTIVATION: Nanopore-based sequencing techniques can reconstruct properties of biosequences by analyzing the sequence-dependent ionic current steps produced as biomolecules pass through a pore. Typically this involves alignment of new data to a reference, where both reference construction and alignment have been performed by hand. RESULTS: We propose an automated method for aligning nanopore data to a reference through the use of hidden Markov models. Several features that arise from prior processing steps and from the class of enzyme used can be simply incorporated into the model. Previously, the M2MspA nanopore was shown to be sensitive enough to distinguish between cytosine, methylcytosine and hydroxymethylcytosine. We validated our automated methodology on a subset of that data by automatically calculating an error rate for the distinction between the three cytosine variants and show that the automated methodology produces a 2-3% error rate, lower than the 10% error rate from previous manual segmentation and alignment. AVAILABILITY AND IMPLEMENTATION: The data, output, scripts and tutorials replicating the analysis are available at https://github.com/UCSCNanopore/Data/tree/master/Automation.

Error rates for nanopore discrimination among cytosine, methylcytosine, and hydroxymethylcytosine along individual DNA strands.

Cytosine, 5-methylcytosine, and 5-hydroxymethylcytosine were identified during translocation of single DNA template strands through a modified Mycobacterium smegmatis porin A (M2MspA) nanopore under control of phi29 DNA polymerase. This identification was based on three consecutive ionic current states that correspond to passage of modified or unmodified CG dinucleotides and their immediate neighbors through the nanopore limiting aperture. To establish quality scores for these calls, we examined ~3,300 translocation events for 48 distinct DNA constructs. Each experiment analyzed a mixture of cytosine-, 5-methylcytosine-, and 5-hydroxymethylcytosine-bearing DNA strands that contained a marker that independently established the correct cytosine methylation status at the target CG of each molecule tested. To calculate error rates for these calls, we established decision boundaries using a variety of machine-learning methods. These error rates depended upon the identity of the bases immediately 5' and 3' of the targeted CG dinucleotide, and ranged from 1.7% to 12.2% for a single-pass read. We estimate that Q40 values (0.01% error rates) for methylation status calls could be achieved by reading single molecules 5-19 times depending upon sequence context.

Nanopores discriminate among five C5-cytosine variants in DNA.

Individual DNA molecules can be read at single nucleotide precision using nanopores coupled to processive enzymes. Discrimination among the four canonical bases has been achieved, as has discrimination among cytosine, 5-methylcytosine (mC), and 5-hydroxymethylcytosine (hmC). Two additional modified cytosine bases, 5-carboxylcytosine (caC) and 5-formylcytosine (fC), are produced during enzymatic conversion of hmC to cytosine in mammalian cells. Thus, an accurate picture of the cytosine epigenetic status in target cells should also include these C5-cytosine variants. In the present study, we used a patch clamp amplifier to acquire ionic current traces caused by phi29 DNA polymerase-controlled translocation of DNA templates through the M2MspA pore. Decision boundaries based on three consecutive ionic current states were implemented to call mC, hmC, caC, fC, or cytosine at CG dinucleotides in ∼4400 individual DNA molecules. We found that the percentage of correct base calls for single pass reads ranged from 91.6% to 98.3%. This accuracy depended upon the identity of nearest neighbor bases surrounding the CG dinucleotide.

Predicting chromatin conformation contact maps.

Over the past 15 years, a variety of next-generation sequencing assays have been developed for measuring the 3D conformation of DNA in the nucleus. Each of these assays gives, for a particular cell or tissue type, a distinct picture of 3D chromatin architecture. Accordingly, making sense of the relationship between genome structure and function requires teasing apart two closely related questions: how does chromatin 3D structure change from one cell type to the next, and how do different measurements of that structure differ from one another, even when the two assays are carried out in the same cell type? In this work, we assemble a collection of chromatin 3D datasets-each represented as a 2D contact map- spanning multiple assay types and cell types. We then build a machine learning model that predicts missing contact maps in this collection. We use the model to systematically explore how genome 3D architecture changes, at the level of compartments, domains, and loops, between cell type and between assay types.

Health Justice and Systems of Care: A Required Longitudinal Course for MD Students.

Problem & Background: Medical education has acknowledged the impact of structural societal factors on health, prompting the need for curricula seeking to eliminate health inequities upstream while simultaneously caring for downstream effects of existing inequities. The Keck School of Medicine of USC (KSOM) implemented one such comprehensive curriculum, Health Justice and Systems of Care (HJSC), integrating health systems science, structural competency, and service-learning in a required course spanning the pre-clerkship and clerkship phases with an optional post clerkship elective. Approach: The HJSC course addresses topics including racism in medicine, health inequities, and health systems science. Using transformative learning theory, it fosters critical consciousness and structural competency. Assessments include case analyses, reflections, team-based learning sessions, and group projects related to social justice in healthcare. The program aims to instill cultural humility and practical application, fostering a holistic approach to medical education that implores physicians to become advocates for health justice. Outcomes of the Innovation: Feedback from students indicated generally positive perceptions of the curriculum. Students provided overall positive comments about discussions with guest speakers. However, students expressed a desire for more concrete examples of how health inequities can be remedied. Some found small-group activities less engaging. Other challenges included providing students of different readiness levels with tailored experiences and seamlessly integrating HJSC content within basic and clinical sciences courses. Critical Reflection: Next steps include continuing to integrate content into the science curriculum and clerkships, improving opportunities for meaningful student interactions, and enhancing faculty development to address health justice concerns in clinical settings.

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.

Mental Health Disparities among Pre-Clinical Medical Students at Saint Louis University during the COVID-19 Pandemic.

The COVID-19 pandemic disproportionately affected racial and ethnic minorities. Medical students were also particularly impacted as they coped with increased stressors due to delayed medical training and a high prevalence of mental health conditions. This study investigates mental health disparities of underrepresented in medicine (URM) students at the Saint Louis University School of Medicine (SLUSOM). An anonymous online survey was distributed to first- and second-year medical students at SLUSOM in February 2021. The survey queried demographic information, lifestyle factors, and pandemic-related and institutional concerns. Mental health was assessed via the Generalized Anxiety Disorder-7 (GAD-7) and the Patient Health Questionnaire-9 (PHQ-9). Statistical tests were run with SPSS, version 27. A convenience sample of 87 students responded to the survey. Students who were categorized as URM were significantly more likely to be at risk of major depressive disorder during the pandemic. Concern about a lack of financial support was significantly greater among students categorized as URM. Concerns regarding a lack of financial support, mental health support, and decreased quality of medical training significantly predicted PHQ-9 scores. Our findings revealed several key factors that may exacerbate mental health disparities among URM students during the pandemic. Providing adequate financial and academic resources for URMs may improve mental health outcomes for similar adverse events in the future.