Lorin A. Crawford

GENESIS predicts single-cell perturbation responses from experimental conditions, outperforming existing models and approaching experimental reproducibility.

The paper examines the role of singleton protein sequences in training protein language models, finding they contain valuable functional diversity and should be included in training.

This study evaluates how pre-training dataset size and diversity affect single-cell foundation model performance on zero-shot and fine-tuned tasks using 22.2 million cells.

This paper introduces EnhancAR, a deep learning model that generates functionally diverse enhancers using evolutionary homologs instead of cell type labels.

This paper introduces a metric to evaluate synergistic integration in multimodal cell models and shows when it benefits tasks over unimodal approaches.

This paper introduces NCLUSION, a nonparametric clustering method for scRNA-seq that identifies marker genes and clusters simultaneously, showing improved performance and reduced runtime.

This study introduces trainable subnetworks to analyze how protein language models factorize structural information and shows that structure prediction is sensitive to these factorized representations.

This paper introduces a pretraining task for genome language models that predicts evolutionary rate, showing improved performance over sequence-only pretraining.

This study examines if synergy scores and toxicity metrics correlate with known adverse drug interactions, finding significant limitations in using them as penalties for toxicity in synergy prediction.

ML-MAGES uses machine learning to improve genetic association analysis by shrinking inflated effect sizes and clustering variant associations across multiple traits.

This paper introduces an adaptive resampling method to improve machine learning performance on imbalanced single-cell datasets, showing better results in gene expression reconstruction, cell type classification, and perturbation response prediction.

The paper introduces SME, a fast method for detecting genome-wide epistasis in large biobank studies by focusing on functionally relevant genomic regions.

This paper introduces a hierarchical cross-entropy loss to improve single-cell annotation models by aligning them with biological hierarchies, achieving better out-of-distribution performance.

This paper introduces a method to prevent over-clustering in scRNA-seq by controlling double-dipping, improving differential expression analysis accuracy.

This study examines how training data composition affects deep learning models in single-cell transcriptomics, finding poor generalization to new cell types and benefits from including embryonic stem cell differentiation data.

This study evaluates machine learning approaches for predicting cancer drug synergy and highlights the need for standardized frameworks in the field.

This study identifies mutation and transcriptional state compatibility in Ph+ ALL MRD, showing that targeting these interactions can overcome BCR::ABL1 independence and improve MRD monitoring.

The paper identifies EBV variants, particularly in LMP1, associated with survival in Burkitt lymphoma and defines prognostic subgroups based on these features.

This paper introduces i-LDSC, an extension of LDSC that accounts for genetic interactions, showing it recovers additional genetic variance in 25 traits from UK Biobank and BioBank Japan.

The paper introduces a probabilistic α-shape sampler to generate realistic 2D and 3D biological shapes from real data, demonstrating its use on neutrophil images and primate mandible CT scans.

Used without fine-tuning, Geneformer and scGPT show limited reliability for cell type clustering and batch effect correction, often landing behind simpler methods.

Multioviz enables in silico perturbation and analysis of gene regulatory networks to guide experimental decisions using a real dataset from mice.

This paper introduces mvMAPIT, a method to detect epistasis in GWAS by jointly modeling multiple traits, showing improved power over single-trait approaches.

This paper introduces a self-supervised learning approach to detect and classify biosynthetic gene clusters in microbial genomes using masked language models trained on protein domains.

This paper introduces a Bayesian neural network approach for variable selection in genetic fine-mapping, showing improved identification of true effect variables in complex traits.

The paper identifies 1,000 gene-level associations and conserved pathway associations for 25 traits across seven ancestries using enrichment analyses on large-scale GWA data.

This paper introduces BANNs, a probabilistic framework for GWA studies that improves association mapping and enrichment analysis by incorporating biological annotations and variational inference.

MegaLMM improves genetic value prediction accuracy by analyzing thousands of traits simultaneously in genomic studies.

This paper introduces SINATRA, a statistical pipeline for identifying physical features differentiating 3D shape classes without requiring correspondences, and demonstrates its application to primate mandibular molars.

The paper identifies genetic modifiers of flight performance in Drosophila using the DGRP, finding additive and epistatic variants, whole genes, and network interactions, with implications for neurodevelopment and neurodegeneration.