publications

2025

1. 

   A DNA language model based on multispecies alignment predicts the effects of genome-wide variants

   Gonzalo Benegas, Carlos Albors, Alan J Aw, Chengzhong Ye, and Yun S Song

   Nature Biotechnology, 2025

   Abs HTML PDF Code

   Protein language models have demonstrated remarkable performance in predicting the effects of missense variants but DNA language models have not yet shown a competitive edge for complex genomes such as that of humans. This limitation is particularly evident when dealing with the vast complexity of noncoding regions that comprise approximately 98% of the human genome. To tackle this challenge, we introduce GPN-MSA (genomic pretrained network with multiple-sequence alignment), a framework that leverages whole-genome alignments across multiple species while taking only a few hours to train. Across several benchmarks on clinical databases (ClinVar, COSMIC and OMIM), experimental functional assays (deep mutational scanning and DepMap) and population genomic data (gnomAD), our model for the human genome achieves outstanding performance on deleteriousness prediction for both coding and noncoding variants. We provide precomputed scores for all ~9 billion possible single-nucleotide variants in the human genome. We anticipate that our advances in genome-wide variant effect prediction will enable more accurate rare disease diagnosis and improve rare variant burden testing.

2. 

   Genomic language models: opportunities and challenges

   Gonzalo Benegas*, Chengzhong Ye*, Carlos Albors*, Jianan Canal Li*, and Yun S Song

   Trends in Genetics, 2025

   Abs HTML PDF

   Large language models (LLMs) are having transformative impacts across a wide range of scientific fields, particularly in the biomedical sciences. Just as the goal of natural language processing is to understand sequences of words, a major objective in biology is to understand biological sequences. Genomic language models (gLMs), which are LLMs trained on DNA sequences, have the potential to significantly advance our understanding of genomes and how DNA elements at various scales interact to give rise to complex functions. To showcase this potential, we highlight key applications of gLMs, including functional constraint prediction, sequence design, and transfer learning. Despite notable recent progress, however, developing effective and efficient gLMs presents numerous challenges, especially for species with large, complex genomes. Here, we discuss major considerations for developing and evaluating gLMs.

2023

1. 

   DNA language models are powerful predictors of genome-wide variant effects

   Gonzalo Benegas, Sanjit Singh Batra, and Yun S. Song

   Proceedings of the National Academy of Sciences, 2023

   Abs HTML PDF Code

   The expanding catalog of genome-wide association studies (GWAS) provides biological insights across a variety of species, but identifying the causal variants behind these associations remains a significant challenge. Experimental validation is both labor-intensive and costly, highlighting the need for accurate, scalable computational methods to predict the effects of genetic variants across the entire genome. Inspired by recent progress in natural language processing, unsupervised pretraining on large protein sequence databases has proven successful in extracting complex information related to proteins. These models showcase their ability to learn variant effects in coding regions using an unsupervised approach. Expanding on this idea, we here introduce the Genomic Pre-trained Network (GPN), a model designed to learn genome-wide variant effects through unsupervised pretraining on genomic DNA sequences. Our model also successfully learns gene structure and DNA motifs without any supervision. To demonstrate its utility, we train GPN on unaligned reference genomes of Arabidopsis thaliana and seven related species within the Brassicales order and evaluate its ability to predict the functional impact of genetic variants in A. thaliana by utilizing allele frequencies from the 1001 Genomes Project and a comprehensive database of GWAS. Notably, GPN outperforms predictors based on popular conservation scores such as phyloP and phastCons. Our predictions for A. thaliana can be visualized as sequence logos in the UCSC Genome Browser (https://genome.ucsc.edu/s/gbenegas/gpn-arabidopsis). We provide code (https://github.com/songlab-cal/gpn) to train GPN for any given species using its DNA sequence alone, enabling unsupervised prediction of variant effects across the entire genome.

2. choros: correction of sequence-based biases for accurate quantification of ribosome profiling data

   Amanda Mok, Robert Tunney, Gonzalo Benegas, Edward W. J. Wallace, and Liana F. Lareau

   bioRxiv, 2023

   Abs HTML

   Ribosome profiling quantifies translation genome-wide by sequencing ribosome-protected fragments, or footprints. Its single-codon resolution allows identification of translation regulation, such as ribosome stalls or pauses, on individual genes. However, enzyme preferences during library preparation lead to pervasive sequence artifacts that obscure translation dynamics. Widespread over- and under-representation of ribosome footprints can dominate local footprint densities and skew estimates of elongation rates by up to five fold. To address these biases and uncover true patterns of translation, we present choros, a computational method that models ribosome footprint distributions to provide bias-corrected footprint counts. choros uses negative binomial regression to accurately estimate two sets of parameters: (i) biological contributions from codon-specific translation elongation rates; and (ii) technical contributions from nuclease digestion and ligation efficiencies. We use these parameter estimates to generate bias correction factors that eliminate sequence artifacts. Applying choros to multiple ribosome profiling datasets, we are able to accurately quantify and attenuate ligation biases to provide more faithful measurements of ribosome distribution. We show that a pattern interpreted as pervasive ribosome pausing near the beginning of coding regions is likely to arise from technical biases. Incorporating choros into standard analysis pipelines will improve biological discovery from measurements of translation.Competing Interest StatementThe authors have declared no competing interest.

2022

1. 

   Robust and annotation-free analysis of alternative splicing across diverse cell types in mice

   Gonzalo Benegas, Jonathan Fischer, and Yun S Song

   eLife, Mar 2022

   Abs HTML Code

   Although alternative splicing is a fundamental and pervasive aspect of gene expression in higher eukaryotes, it is often omitted from single-cell studies due to quantification challenges inherent to commonly used short-read sequencing technologies. Here, we undertake the analysis of alternative splicing across numerous diverse murine cell types from two large-scale single-cell datasets—the \textitTabula Muris and BRAIN Initiative Cell Census Network—while accounting for understudied technical artifacts and unannotated events. We find strong and general cell-type-specific alternative splicing, complementary to total gene expression but of similar discriminatory value, and identify a large volume of novel splicing events. We specifically highlight splicing variation across different cell types in primary motor cortex neurons, bone marrow B cells, and various epithelial cells, and we show that the implicated transcripts include many genes which do not display total expression differences. To elucidate the regulation of alternative splicing, we build a custom predictive model based on splicing factor activity, recovering several known interactions while generating new hypotheses, including potential regulatory roles for novel alternative splicing events in critical genes like \textitKhdrbs3 and \textitRbfox1. We make our results available using public interactive browsers to spur further exploration by the community.