Biology subjects

Gitter, A.

Publications and source records attributed to Gitter, A..

At least 19 recordsLinked to original sources

bioRxiv2026

A Framework for Benchmarking Pathway Reconstruction Algorithms

Cells coordinate diverse biological processes through interactions among thousands of molecules, but mapping these interactions comprehensively and systematically remains an open problem. Pathway reconstruction algorithms address this problem by linking molecules of interest, identified from high-throughput omics experiments, using prior knowledge encoded as background interaction networks. This process recovers intermediate molecules and interactions that were not directly measured in the experimental data but plausibly connect the observed molecules. It generates testable hypotheses about interactions that drive cell behavior and informs the choice of follow-up experiments. Many algorithms have been created over decades, each optimizing different computational objectives and relying on different assumptions. The resulting heterogeneity has made benchmarking challenging, limiting systematic comparisons. Therefore, selecting an algorithm for a given biological context remains a non-trivial and poorly informed task. This registered report presents a large-scale benchmark of pathway reconstruction algorithms, evaluating 14 algorithms across 822 datasets from four biological settings. To enable this benchmark, we introduce Signaling Pathway Reconstruction Analysis Streamliner (SPRAS), which standardizes algorithm inputs, outputs, and execution into a formal framework, enabling systematic comparison that was previously infeasible. We will assess each algorithm on reconstruction performance against gold standard pathways, algorithm similarity, and computational performance across different biological contexts. Together, these evaluations will provide quantitative evidence for understanding pathway reconstruction algorithm behavior and guiding algorithm selection.

bioinformatics↗

bioRxiv2026

BoltzMol-1: Towards Reliable Virtual Screening for Fast and Cost-Effective Hit Discovery

We present BoltzMol-1, a small-molecule hit discovery pipeline, centered on an optimized version of Boltz-2, explicitly adapted for prospective discovery. Reliable hit discovery that generalizes across target classes (rather than only the well-characterized families that dominate existing ligand data) would broaden the range of biology accessible to small-molecule intervention and reduce reliance on resource-intensive high-throughput screening. Towards this goal, the system prioritizes compounds for rapid experimental validation by coupling model-driven ranking with streamlined procurement from commercial catalogs. To improve developability at the point of selection, we introduce a suite of ADMET models for kinetic solubility (logS), lipophilicity (logD), and Caco-2 permeability. These models act as an early triage layer, systematically filtering out compounds with unfavorable physicochemical and absorption properties prior to synthesis or purchase. Across a panel of ten targets (most with no representation in the underlying affinity training data) we observe strong prospective performance on challenging systems. Functional actives or binders were identified for 6 of 10 targets, despite modest experimental budgets of 28-96 compounds per target. These results include successes on receptors and enzymes traditionally considered difficult for structure- or ligand-based approaches. Collectively, this work establishes a practical framework for low-throughput, cost-constrained discovery campaigns capable of delivering chemically tractable binders with favorable property profiles. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=124 SRC="FIGDIR/small/736485v1_fig1.gif" ALT="Figure 1"> View larger version (50K): org.highwire.dtl.DTLVardef@1d7ff06org.highwire.dtl.DTLVardef@1a7fc7aorg.highwire.dtl.DTLVardef@1b0dc98org.highwire.dtl.DTLVardef@62b978_HPS_FORMAT_FIGEXP M_FIG O_FLOATNOFigure 1:C_FLOATNO Overview of the prospective virtual-screening campaigns across all targets. For each target, the panel shows the predicted protein-ligand complex together with the number of compounds tested, the number of confirmed actives/binders, and the assays used for screening and follow-up. C_FIG

biochemistry↗

bioRxiv2025

Crowdsourced Protein Design: Lessons From the Adaptyv EGFR Binder Competition

In this report, we summarize and analyze the 2024 Adaptyv protein design competition. Participants used computational and Machine Learning (ML) methods of their choice to design proteins that bind the Epidermal Growth Factor Receptor (EGFR), a key drug target involved in cell growth, differentiation, and cancer development. Over 1,800 designs were submitted across two rounds. Of these, 601 proteins were selected and characterized for expression and binding affinity to EGFR, with competitors both optimizing existing binders (KD = 1.21 nM) and creating de novo binders (KD = 82 nM). All selected designs were experimentally validated using Adaptyvs automated Bio-Layer Interferometry (BLI) pipeline. This competition illustrates the potential of crowdsourcing to drive creativity and innovation in protein design. However, it also exposed key challenges, such as the lack of standardized benchmarks, experimental design targets, and robust computational metrics for method comparison. We anticipate that future competitions will address these gaps and further motivate progress in computational protein design.

bioengineering↗

bioRxiv2024

Protein Set Transformer: A protein-based genome language model to power high diversity viromics

Exponential increases in microbial and viral genomic data demand transformational advances in scalable, generalizable frameworks for their interpretation. Standard homology-based functional analyses are hindered by the rapid divergence of microbial and especially viral genomes and proteins that significantly decreases the volume of usable data. Here, we present Protein Set Transformer (PST), a protein-based genome language model that models genomes as sets of proteins without considering sparsely available functional labels. Trained on >100k viruses, PST outperformed other homology- and language model-based approaches for relating viral genomes based on shared protein content. Further, PST demonstrated protein structural and functional awareness by clustering capsid-fold-containing proteins with known capsid proteins and uniquely clustering late gene proteins within related viruses. Our data establish PST as a valuable method for diverse viral genomics, ecology, and evolutionary applications. We posit that the PST framework can be a foundation model for microbial genomics when trained on suitable data.

bioinformatics↗

bioRxiv2024

MPAC: a computational framework for inferring cancer pathway activities from multi-omic data

Fully capturing cellular state requires examining genomic, epigenomic, transcriptomic, proteomic, and other assays for a biological sample and comprehensive computational modeling to reason with the complex and sometimes conflicting measurements. Modeling these so-called multi-omic data is especially beneficial in disease analysis, where observations across omic data types may reveal unexpected patient groupings and inform clinical outcomes and treatments. We present Multi-omic Pathway Analysis of Cells (MPAC), a computational framework that interprets multi-omic data through prior knowledge from biological pathways. MPAC leverages network relationships encoded in pathways through a factor graph to infer consensus activity levels for proteins and associated pathway entities from multi-omic data, runs permutation testing to eliminate spurious activity predictions, and groups biological samples by pathway activities to allow identifying and prioritizing proteins with potential clinical relevance, e.g., associated with patient prognosis. Using DNA copy number alteration and RNA-seq data from head and neck squamous cell carcinoma patients from The Cancer Genome Atlas as an example, we demonstrate that MPAC predicts a patient subgroup related to immune responses not identified by analysis with either input omic data type alone. Key proteins identified via this subgroup have pathway activities related to clinical outcome as well as immune cell compositions. Our MPAC R package, available at https://bioconductor.org/packages/MPAC, enables similar multi-omic analyses on new datasets.

bioinformatics↗

bioRxiv2024

Biophysics-based protein language models for protein engineering

Protein language models trained on evolutionary data have emerged as powerful tools for predictive problems involving protein sequence, structure, and function. However, these models overlook decades of research into biophysical factors governing protein function. We propose Mutational Effect Transfer Learning (METL), a protein language model framework that unites advanced machine learning and biophysical modeling. Using the METL framework, we pretrain transformer-based neural networks on biophysical simulation data to capture fundamental relationships between protein sequence, structure, and energetics. We finetune METL on experimental sequence-function data to harness these biophysical signals and apply them when predicting protein properties like thermostability, catalytic activity, and fluorescence. METL excels in challenging protein engineering tasks like generalizing from small training sets and position extrapolation, although existing methods that train on evolutionary signals remain powerful for many types of experimental assays. We demonstrate METLs ability to design functional green fluorescent protein variants when trained on only 64 examples, showcasing the potential of biophysics-based protein language models for protein engineering.

bioinformatics↗

bioRxiv2022

An approachable, flexible, and practical machine learning workshop for biologists

The increasing prevalence and importance of machine learning in biological research has created a need for machine learning training resources tailored towards biological researchers. However, existing resources are often inaccessible, infeasible, or inappropriate for biologists because they require significant computational and mathematical knowledge, demand an unrealistic time-investment, or teach skills primarily for computational researchers. We created the Machine Learning for Biologists (ML4Bio) workshop, a short, intensive workshop that empowers biological researchers to comprehend machine learning applications and pursue machine learning collaborations in their own research. The ML4Bio workshop focuses on classification and was designed around 3 principles: (a) focusing on preparedness over fluency or expertise, (b) necessitating minimal coding and mathematical background, and (c) requiring low time investment. It incorporates active learning methods and custom open source software that allows participants to explore machine learning workflows. After multiple sessions to improve workshop design, we performed a study on 3 workshop sessions. Despite some confusion around identifying subtle methodological flaws in machine learning workflows, participants generally reported that the workshop met their goals, provided them with valuable skills and knowledge, and greatly increased their beliefs that they could engage in research that uses machine learning. ML4Bio is an educational tool for biological researchers, and its creation and evaluation provides valuable insight into tailoring educational resources for active researchers in different domains. The workshop materials are available from https://carpentries-incubator.github.io/ml4bio-workshop/ and the ml4bio software is available from https://github.com/gitter-lab/ml4bio.

scientific communication and education↗

bioRxiv2020

Alternative splicing liberates a cryptic cytoplasmic isoform of mitochondrial MECR that antagonizes influenza virus

Viruses must balance their reliance on host cell machinery for replication while avoiding host defense. Influenza A viruses are zoonotic agents that frequently switch hosts, causing localized outbreaks with the potential for larger pandemics. The host range of influenza virus is limited by the need for successful interactions between the virus and cellular partners. Here we used immuno-competitive capture-mass spectrometry to identify cellular proteins that interact with human- and avian-style viral polymerases. We focused on the pro-viral activity of heterogenous nuclear ribonuclear protein U-like 1 (hnRNP UL1) and the anti-viral activity of mitochondrial enoyl CoA-reductase (MECR). MECR is localized to mitochondria where it functions in mitochondrial fatty acid synthesis (mtFAS). While a small fraction of the polymerase subunit PB2 localizes to the mitochondria, PB2 did not interact with full-length MECR. By contrast, a minor splice variant produces cytoplasmic MECR (cMECR) that interacts with PB2. cMECR binds the viral polymerase and suppresses viral replication by blocking assembly of viral ribonucleoprotein complexes (RNPs). MECR ablation through genome editing or drug treatment is detrimental for cell health, creating a generic block to virus replication. Using the yeast homolog Etr1 to supply the metabolic functions of MECR in MECR-null cells, we showed that specific antiviral activity is independent of mtFAS and lies solely within cMECR. Thus, alternative splicing produces a cryptic antiviral protein that is embedded within a key metabolic enzyme.

microbiology↗

bioRxiv2020

Neural networks to learn protein sequence-function relationships from deep mutational scanning data

The mapping from protein sequence to function is highly complex, making it challenging to predict how sequence changes will affect a proteins behavior and properties. We present a supervised deep learning framework to learn the sequence-function mapping from deep mutational scanning data and make predictions for new, uncharacterized sequence variants. We test multiple neural network architectures, including a graph convolutional network that incorporates protein structure, to explore how a networks internal representation affects its ability to learn the sequence-function mapping. Our supervised learning approach displays superior performance over physics-based and unsupervised prediction methods. We find networks that capture nonlinear interactions and share parameters across sequence positions are important for learning the relationship between sequence and function. Further analysis of the trained models reveals the networks ability to learn biologically meaningful information about protein structure and mechanism. Finally, we demonstrate the models ability to navigate sequence space and design new proteins beyond the training set. We applied the GB1 models to design a sequence that binds to IgG with substantially higher affinity than wild-type GB1. Our software is available from https://github.com/gitter-lab/nn4dms. SignificanceUnderstanding the relationship between protein sequence and function is necessary to design new and useful proteins with applications in bioenergy, medicine, and agriculture. The mapping from sequence to function is tremendously complex because it involves thousands of molecular interactions that are coupled over multiple lengths and timescales. In this work, we show neural networks can learn the sequence-function mapping from large protein datasets. Neural networks are appealing for this task because they can learn complicated relationships from data, make few assumptions about the nature of the sequencefunction relationship, and can learn general rules that apply across the length of the protein sequence. We demonstrate the learned models can be applied to design new proteins with properties that exceed natural sequences.

bioinformatics↗

bioRxiv2019

Automating parameter selection to avoid implausible biological pathway models

A common way to integrate and analyze large amounts of biological "omic" data is through pathway reconstruction: using condition-specific omic data to create a subnetwork of a generic background network that represents some process or cellular state. A challenge in pathway reconstruction is that adjusting pathway reconstruction algorithms parameters produces pathways with drastically different topological properties and biological interpretations. Due to the exploratory nature of pathway reconstruction, there is no ground truth for direct evaluation, so parameter tuning methods typically used in statistics and machine learning are inapplicable. We developed the pathway parameter advising algorithm to tune pathway reconstruction algorithms to minimize biologically implausible predictions. We leverage background knowledge in pathway databases to select pathways whose high-level structure resembles that of manually curated biological pathways. At the core of this method is a graphlet decomposition metric, which measures topological similarity to curated biological pathways. In order to evaluate pathway parameter advising, we compare its performance in avoiding implausible networks and reconstructing pathways from the NetPath database with other parameter selection methods across four pathway reconstruction algorithms. We also demonstrate how pathway parameter advising can guide construction of an influenza host factor network. Pathway parameter advising is method-agnostic; it is applicable to any pathway reconstruction algorithm with tunable parameters. Our pathway parameter advising software is available on GitHub at https://github.com/gitter-lab/pathway-parameter-advising and PyPI at https://pypi.org/project/pathwayParameterAdvising/.

bioinformatics↗

bioRxiv2019

Classifying T cell activity in autofluorescence intensity images with convolutional neural networks

The importance of T cells in immunotherapy has motivated developing technologies to better characterize T cells and improve therapeutic efficacy. One specific objective is assessing antigen-induced T cell activation because only functionally active T cells are capable of killing the desired targets. Autofluorescence imaging can distinguish T cell activity states of individual cells in a non-destructive manner by detecting endogenous changes in metabolic co-enzymes such as NAD(P)H. However, recognizing robust patterns of T cell activity is computationally challenging in the absence of exogenous labels or information-rich autofluorescence lifetime measurements. We demonstrate that advanced machine learning can accurately classify T cell activity from NAD(P)H intensity images and that those image-based signatures transfer across human donors. Using a dataset of 8,260 cropped single-cell images from six donors, we meticulously evaluate multiple machine learning models. These range from traditional models that represent images using summary statistics or extract image features with CellProfiler to deep convolutional neural networks (CNNs) pre-trained on general non-biological images. Adapting pre-trained CNNs for the T cell activity classification task provides substantially better performance than traditional models or a simple CNN trained with the autofluorescence images alone. Visualizing the images with dimension reduction provides intuition into why the CNNs achieve higher accuracy than other approaches. However, we observe that fine-tuning all layers of the pre-trained CNN does not provide a classification performance boost commensurate with the additional computational cost. Our software detailing our image processing and model training pipeline is available as Jupyter notebooks at https://github.com/gitter-lab/t-cell-classification.

bioinformatics↗

bioRxiv2019

Network Inference with Granger Causality Ensembles on Single-Cell Transcriptomic Data

Advances in single-cell transcriptomics enable measuring the gene expression of individual cells, allowing cells to be ordered by their state in a dynamic biological process. Many algorithms assign pseudotimes to each cell, representing the progress along the biological process. Ordering the expression data according to such pseudotimes can be valuable for understanding the underlying regulator-gene interactions in a biological process, such as differentiation. However, the distribution of cells sampled along a transitional process, and hence that of the pseudotimes assigned to them, is not uniform. This prevents using many standard mathematical methods for analyzing the ordered gene expression states. We present Single-cell Inference of Networks using Granger Ensembles (SINGE), an algorithm for gene regulatory network inference from single-cell gene expression data. Given ordered single-cell data, SINGE uses kernel-based Granger Causality regression, which smooths the irregular pseudotimes and missing expression values. It then aggregates the predictions from an ensemble of regression analyses with a modified Borda count to compile a ranked list of candidate interactions between transcriptional regulators and their target genes. In two mouse embryonic stem cell differentiation case studies, SINGE outperforms other contemporary algorithms for gene network reconstruction. However, a more detailed examination reveals caveats about transcriptional network reconstruction with single-cell RNA-seq data. Network inference methods, including SINGE, may have near random performance for predicting the targets of many individual regulators even if the overall performance is good. In addition, including uninformative pseudotime values can hurt the performance of network reconstruction methods. A MATLAB implementation of SINGE is available at https://github.com/gitter-lab/SINGE.

bioinformatics↗

bioRxiv2019

Predicting kinase inhibitors using bioactivity matrix derived informer sets

Prediction of compounds that are active against a desired biological target is a common step in drug discovery efforts. Virtual screening methods seek some active-enriched fraction of a library for experimental testing. Where data are too scarce to train supervised learning models for compound prioritization, initial screening must provide the necessary data. Commonly, such an initial library is selected on the basis of chemical diversity by some pseudo-random process (for example, the first few plates of a larger library) or by selecting an entire smaller library. These approaches may not produce a sufficient number or diversity of actives. An alternative approach is to select an informer set of screening compounds on the basis of chemogenomic information from previous testing of compounds against a large number of targets. We compare different ways of using chemogenomic data to choose a small informer set of compounds based on previously measured bioactivity data. We develop this Informer-Based-Ranking (IBR) approach using the Published Kinase Inhibitor Sets (PKIS) as the chemogenomic data to select the informer sets. We test the informer compounds on a target that is not part of the chemogenomic data, then predict the activity of the remaining compounds based on the experimental informer data and the chemogenomic data. Through new chemical screening experiments, we demonstrate the utility of IBR strategies in a prospective test on two kinase targets not included in the PKIS. Using limited training data in both retrospective and prospective tests, bioactivity fingerprints based on chemogenomic data outperform chemical fingerprints in predicting active compounds in both standard virtual screening metrics and accurate identification of hits from novel chemical classes. Author SummaryIn the early stages of drug discovery efforts, computational models are used to predict activity and prioritize compounds for experimental testing. New targets commonly lack the data necessary to build effective models, and the screening needed to generate that experimental data can be costly. We seek to improve the efficiency of the initial screening phase, and of the process of prioritizing compounds for subsequent screening. We choose a small informer set of compounds based on publicly available prior screening data on distinct (though related) targets. We then use experimental data on these informer compounds to predict the activity of other compounds in the set against the target of interest. Computational and statistical tools are needed to identify informer compounds and to prioritize other compounds for subsequent phases of screening. Using limited training data, we find that selection of informer compounds on the basis of bioactivity data from previous screening efforts is superior to the traditional approach of selection of a chemically diverse subset of compounds. We demonstrate the success of this approach in retrospective tests on the Published Kinase Inhibitor Sets (PKIS) chemogenomic data and in prospective experimental screens against two additional non-human kinase targets.

pharmacology and toxicology↗

bioRxiv2018

Learning Molecule Drug Function from Structure Representations with Deep Neural Networks or Random Forests

Empirical testing of chemicals for desirable properties, such as drug efficacy, toxicity, solubility, and thermal conductivity, costs many billions of dollars every year. Further, the choice of molecules for testing relies on expert knowledge and intuition in the relevant domain of application. The ability to predict the action of a molecule in silico would greatly increase the speed and decrease the cost of prioritizing molecules with desirable function for experimental testing. We explore the use of molecule structures to predict their drug classification without any explicit biological model (e.g. protein structure or cell system). Two approaches were compared: (1) a two-dimensional image representation of molecule structure with transfer learning from a pre-trained convolutional neural network (CNN), and (2) Morgan molecular fingerprints (MFP) with a random forest (RF). Both methods using only molecule information as input achieved significantly better accuracy than previous work that used transcriptomic measurements of molecular effects. Because these data suggest that much of a molecules function is encoded by its chemical structure, we explore which general molecular properties are drug class-specific and how misclassification of structures might provide drug repurposing opportunities.

bioinformatics↗

bioRxiv2018

Practical model selection for prospective virtual screening

Virtual (computational) high-throughput screening provides a strategy for prioritizing compounds for experimental screens, but the choice of virtual screening algorithm depends on the dataset and evaluation strategy. We consider a wide range of ligand-based machine learning and docking-based approaches for virtual screening on two protein-protein interactions, PriA-SSB and RMI-FANCM, and present a strategy for choosing which algorithm is best for prospective compound prioritization. Our workflow identifies a random forest as the best algorithm for these targets over more sophisticated neural network-based models. The top 250 predictions from our selected random forest recover 37 of the 54 active compounds from a library of 22,434 new molecules assayed on PriA-SSB. We show that virtual screening methods that perform well in public datasets and synthetic benchmarks, like multi-task neural networks, may not always translate to prospective screening performance on a specific assay of interest.

biochemistry↗

bioRxiv2018

Lag Penalized Weighted Correlation for Time Series Clustering

MotivationThe similarity or distance measure used for clustering can generate intuitive and interpretable clusters when it is tailored to the unique characteristics of the data. In time series datasets, measurements such as gene expression levels or protein phosphorylation intensities are collected sequentially over time, and the similarity score should capture this special temporal structure.\n\nResultsWe propose a clustering similarity measure called Lag Penalized Weighted Correlation (LPWC) to group pairs of time series that exhibit closely-related behaviors over time, even if the timing is not perfectly synchronized. LPWC aligns pairs of time series profiles to identify common temporal patterns. It down-weights aligned profiles based on the length of the temporal lags that are introduced. We demonstrate the advantages of LPWC versus existing time series and general clustering algorithms. In a simulated dataset based on the biologically-motivated impulse model, LPWC is the only method to recover the true clusters for almost all simulated genes. LPWC also identifies distinct temporal patterns in our yeast osmotic stress response and axolotl limb regeneration case studies.\n\nAvailabilityThe LPWC R package is available at https://github.com/gitter-lab/LPWC and CRAN under a MIT license.\n\nContactchandereng@wisc.edu or gitter@biostat.wisc.edu\n\nSupplementary informationSupplementary files are available online.

bioinformatics↗

bioRxiv2017

Synthesizing Signaling Pathways from Temporal Phosphoproteomic Data

Advances in proteomics reveal that pathway databases fail to capture the majority of cellular signaling activity. Our mass spectrometry study of the dynamic epidermal growth factor (EGF) response demonstrates that over 89% of significantly (de)phosphorylated proteins are excluded from individual EGF signaling maps, and 63% are absent from all annotated pathways. We present a computational method, the Temporal Pathway Synthesizer (TPS), to discover missing pathway elements by modeling temporal phosphoproteomic data. TPS uses constraint solving to exhaustively explore all possible structures for a signaling pathway, eliminating structures that are inconsistent with protein-protein interactions or the observed phosphorylation event timing. Applied to our EGF response data, TPS connects 83% of the responding proteins to receptors and signaling proteins in EGF pathway maps. Inhibiting predicted active kinases supports the TPS pathway model. The TPS algorithm is broadly applicable and also recovers an accurate model of the yeast osmotic stress response.

bioinformatics↗

bioRxiv2017

Network inference reveals novel connections in pathways regulating growth and defense in the yeast salt response

Cells respond to stressful conditions by coordinating a complex, multi-faceted response that spans many levels of physiology. Much of the response is coordinated by changes in protein phosphorylation. Although the regulators of transcriptome changes during stress are well characterized in Saccharomyces cerevisiae, the upstream regulatory network controlling protein phosphorylation is less well dissected. Here, we developed a computational approach to infer the signaling network that regulates phosphorylation changes in response to salt stress. The method uses integer linear programming (ILP) to integrate stress-responsive phospho-proteome responses in wild-type and mutant strains, predicted phosphorylation motifs on groups of coregulated peptides, and published protein interaction data. A key advance is that by grouping peptides into submodules before inference, the method can overcome missing protein interactions in published datasets to predict novel, stress-dependent protein interactions and phosphorylation events. The network we inferred predicted new regulatory connections between stress-activated and growth-regulating pathways and suggested mechanisms coordinating metabolism, cell-cycle progression, and growth during stress. We confirmed several network predictions with co-immunoprecipitations coupled with mass-spectrometry protein identification and mutant phospho-proteomic analysis. Results show that the cAMP-phosphodiesterase Pde2 physically interacts with many stress-regulated transcription factors targeted by PKA, and that reduced phosphorylation of those factors during stress requires the Rck2 kinase that we show physically interacts with Pde2. Together, our work shows how a high-quality computational network model can facilitate discovery of new pathway interactions during osmotic stress.

systems biology↗