Biology subjects

Song, Y. S.

Publications and source records attributed to Song, Y. S..

At least 55 records · Page 3Linked to original sources

bioRxiv2021

Longitudinal single-cell epitope and RNA-sequencing reveals the immunological impact of type 1 interferon autoantibodies in critical COVID-19

Type I interferon (IFN-I) neutralizing autoantibodies have been found in some critical COVID-19 patients; however, their prevalence and longitudinal dynamics across the disease severity scale, and functional effects on circulating leukocytes remain unknown. Here, in 284 COVID-19 patients, we found IFN-I autoantibodies in 19% of critical, 6% of severe and none of the moderate cases. Longitudinal profiling of over 600,000 peripheral blood mononuclear cells using multiplexed single-cell epitope and transcriptome sequencing from 54 COVID-19 patients, 15 non-COVID-19 patients and 11 non-hospitalized healthy controls, revealed a lack of IFN-I stimulated gene (ISG-I) response in myeloid cells from critical cases, including those producing anti-IFN-I autoantibodies. Moreover, surface protein analysis showed an inverse correlation of the inhibitory receptor LAIR-1 with ISG-I expression response early in the disease course. This aberrant ISG-I response in critical patients with and without IFN-I autoantibodies, supports a unifying model for disease pathogenesis involving ISG-I suppression via convergent mechanisms.

immunology↗

bioRxiv2020

The distribution of waiting distances in ancestral recombination graphs and its applications

The ancestral recombination graph (ARG) contains the full genealogical information of the sample, and many population genetic inference problems can be solved using inferred or sampled ARGs. In particular, the waiting distance between tree changes along the genome can be used to make inference about the distribution and evolution of recombination rates. To this end, we here derive an analytic expression for the distribution of waiting distances between tree changes under the sequentially Markovian coalescent model and obtain an accurate approximation to the distribution of waiting distances for topology changes. We use these results to show that some of the recently proposed methods for inferring sequences of trees along the genome provide strongly biased distributions of waiting distances. In addition, we provide a correction to an undercounting problem facing all available ARG inference methods, thereby facilitating the use of ARG inference methods to estimate temporal changes in the recombination rate.

evolutionary biology↗

bioRxiv2020

Single Layers of Attention Suffice to Predict Protein Contacts

AO_SCPLOWBSTRACTC_SCPLOWThe established approach to unsupervised protein contact prediction estimates co-evolving positions using undirected graphical models. This approach trains a Potts model on a Multiple Sequence Alignment, then predicts that the edges with highest weight correspond to contacts in the 3D structure. On the other hand, increasingly large Transformers are being pretrained on protein sequence databases but have demonstrated mixed results for downstream tasks, including contact prediction. This has sparked discussion about the role of scale and attention-based models in unsupervised protein representation learning. We argue that attention is a principled model of protein interactions, grounded in real properties of protein family data. We introduce a simplified attention layer, factored attention, and show that it achieves comparable performance to Potts models, while sharing parameters both within and across families. Further, we extract contacts from the attention maps of a pretrained Transformer and show they perform competitively with the other two approaches. This provides evidence that large-scale pretraining can learn meaningful protein features when presented with unlabeled and unaligned data. We contrast factored attention with the Transformer to indicate that the Transformer leverages hierarchical signal in protein family databases not captured by our single-layer models. This raises the exciting possibility for the development of powerful structured models of protein family databases.1

bioinformatics↗

bioRxiv2020

RNA-Sieve: A likelihood-based deconvolution of bulk gene expression data using single-cell references

Direct comparison of bulk gene expression profiles is complicated by distinct cell type mixtures in each sample which obscure whether observed differences are actually due to changes in expression levels themselves or simply due to differing cell type compositions. Single-cell technology has made it possible to measure gene expression in individual cells, achieving higher resolution at the expense of increased noise. If carefully incorporated, such single-cell data can be used to deconvolve bulk samples to yield accurate estimates of the true cell type proportions, thus enabling one to disentangle the effects of differential expression and cell type mixtures. Here, we propose a generative model and a likelihood-based inference method that uses asymptotic statistical theory and a novel optimization procedure to perform deconvolution of bulk RNA-seq data to produce accurate cell type proportion estimates. We demonstrate the effectiveness of our method, called RNA-Sieve, across a diverse array of scenarios involving real data and discuss extensions made uniquely possible by our probabilistic framework, including a demonstration of well-calibrated confidence intervals.

genomics↗

bioRxiv2020

Transporter genes in biosynthetic gene clusters predict metabolite characteristics and siderophore activity

Biosynthetic gene clusters (BGCs) are operonic sets of microbial genes that synthesize specialized metabolites with diverse functions, including siderophores and antibiotics, which often require export to the extracellular environment. For this reason, genes for transport across cellular membranes are essential for the production of specialized metabolites, and are often genomically co-localized with BGCs. Here we conducted a comprehensive computational analysis of transporters associated with characterized BGCs. In addition to known exporters, in BGCs we found many importer-specific transmembrane domains that co-occur with substrate binding proteins possibly for uptake of siderophores or metabolic precursors. Machine learning models using transporter gene frequencies were predictive of known siderophore activity, molecular weights, and a measure of lipophilicity (log P) for corresponding BGC-synthesized metabolites. Transporter genes associated with BGCs were often equally or more predictive of metabolite features than biosynthetic genes. Given the importance of siderophores as pathogenicity factors, we used transporters specific for siderophore BGCs to identify both known and uncharacterized siderophore-like BGCs in genomes from metagenomes from the infant and adult gut microbiome. We find that 23% of microbial genomes from the infant gut have siderophore-like BGCs, but only 3% of those assembled from adult gut microbiomes do. While siderophore-like BGCs from the infant gut are predominantly associated with Enterobactericaee and Staphylococcus, siderophore-like BGCs can be identified from taxa in the adult gut microbiome that have rarely been recognized for siderophore production. Taken together, these results show that consideration of BGC-associated transporter genes can inform predictions of specialized metabolite structure and function.

bioinformatics↗

bioRxiv2020

SCITO-seq: single-cell combinatorial indexed cytometry sequencing

The development of DNA-barcoded antibodies to tag cell-surface molecules has enabled the use of droplet-based single cell sequencing (dsc-seq) to profile the surface proteomes of cells. Compared to flow and mass cytometry, the major limitation of current dsc-seq-based workflows is the high cost associated with profiling each cell, thus precluding its use in applications where millions of cells are required. Here, we introduce SCITO-seq, a new workflow that combines combinatorial indexing and commercially available dsc-seq to enable cost-effective cell surface proteomic sequencing of greater than 105 cells per microfluidic reaction. We demonstrate SCITO-seqs feasibility and scalability by profiling mixed species cell lines and mixed human T and B lymphocytes. To further demonstrate its applicability, we show comparable cellular composition estimates in peripheral blood mononuclear cells obtained with SCITO-seq and mass cytometry. SCITO-seq can be extended to include simultaneous profiling of additional modalities such as transcripts and accessible chromatin or tracking of experimental perturbations such as genome edits or extracellular stimuli.

genomics↗

bioRxiv2019

Cytoplasmic mRNA decay factors modulate RNA polymerase II processivity in 5’ and 3’ gene regions in yeast

mRNA levels are determined by the balance between mRNA synthesis and decay. Factors that mediate both processes, including the 5 to 3 exonuclease Xrn1, are responsible for the cross talk between the two processes in a manner that buffers steady-state mRNA levels. However, these proteins roles in transcription remain elusive and controversial. Applying NET-seq to yeast cells, we show that Xrn1 functions mainly as a transcriptional activator and that its disruption manifests via the reduction of RNA polymerase II (Pol II) occupancy downstream of transcription start sites. We combine our data and novel mathematical modeling of transcription to suggest that transcription initiation and elongation of targeted genes is modulated by Xrn1. Furthermore, Pol II occupancy markedly increases near cleavage and polyadenylation sites in xrn1{Delta} cells while its activity decreases, a characteristic feature of backtracked Pol II. We also provide indirect evidence that Xrn1 is involved in transcription termination downstream of polyadenylation sites. Two additional decay factors, Dhh1 and Lsm1, seem to function similarly to Xrn1 in transcription, perhaps as a complex, while the decay factors Ccr4 and Rpb4 also perturb transcription in other ways. Interestingly, DFs are capable of differentiating between SAGA- and TFIID-dominated promoters. These two classes of genes respond differently to XRN1 deletion in mRNA synthesis and differentially utilize mRNA decay pathways, raising the possibility that one distinction between the two types of genes lies in the mechanism(s) that balance these processes.

genomics↗

bioRxiv2019

Accurate assembly of the olive baboon (Papio anubis) genome using long-read and Hi-C data

Besides macaques, baboons are the most commonly used nonhuman primate in biomedical research. Despite this importance, the genomic resources for baboons are quite limited. In particular, the current baboon reference genome Panu_3.0 is a highly fragmented, reference-guided (i.e., not fully de novo) assembly, and its poor quality inhibits our ability to conduct downstream genomic analyses. Here we present a truly de novo genome assembly of the olive baboon (Papio anubis) that uses data from several recently developed single-molecule technologies. Our assembly, Panubis1.0, has an N50 contig size of ~1.46 Mb (as opposed to 139 Kb for Panu_3.0), and has single scaffolds that span each of the 20 autosomes and the X chromosome. We highlight multiple lines of evidence (including Bionano Genomics data, pedigree linkage information, and linkage disequilibrium data) suggesting that there are several large assembly errors in Panu_3.0, which have been corrected in Panubis1.0.

genomics↗

bioRxiv2019

Evaluating Protein Transfer Learning with TAPE

Protein modeling is an increasingly popular area of machine learning research. Semi-supervised learning has emerged as an important paradigm in protein modeling due to the high cost of acquiring supervised protein labels, but the current literature is fragmented when it comes to datasets and standardized evaluation techniques. To facilitate progress in this field, we introduce the Tasks Assessing Protein Embeddings (TAPE), a set of five biologically relevant semi-supervised learning tasks spread across different domains of protein biology. We curate tasks into specific training, validation, and test splits to ensure that each task tests biologically relevant generalization that transfers to real-life scenarios. We bench-mark a range of approaches to semi-supervised protein representation learning, which span recent work as well as canonical sequence learning techniques. We find that self-supervised pretraining is helpful for almost all models on all tasks, more than doubling performance in some cases. Despite this increase, in several cases features learned by self-supervised pretraining still lag behind features extracted by state-of-the-art non-neural techniques. This gap in performance suggests a huge opportunity for innovative architecture design and improved modeling paradigms that better capture the signal in biological sequences. TAPE will help the machine learning community focus effort on scientifically relevant problems. Toward this end, all data and code used to run these experiments are available at https://github.com/songlab-cal/tape.

synthetic biology↗

bioRxiv2019

Necrotizing enterocolitis is preceded by increased gut bacterial replication, Klebsiella, and fimbriae-encoding bacteria that may stimulate TLR4 receptors

Necrotizing enterocolitis (NEC) is a devastating intestinal disease that occurs primarily in premature infants. We performed genome-resolved metagenomic analysis of 1,163 fecal samples from premature infants to identify microbial features predictive of NEC. Features considered include genes, bacterial strain types, eukaryotes, bacteriophages, plasmids and growth rates. A machine learning classifier found that samples collected prior to NEC diagnosis harbored significantly more Klebsiella, bacteria encoding fimbriae, and bacteria encoding secondary metabolite gene clusters related to quorum sensing and bacteriocin production. Notably, replication rates of all bacteria, especially Enterobacteriaceae, were significantly higher two days before NEC diagnosis. The findings uncover biomarkers that could lead to early detection of NEC and targets for microbiome-based therapeutics.

microbiology↗

bioRxiv2019

Inference and analysis of population-specific fine-scale recombination maps across 26 diverse human populations

Fine-scale rates of meiotic recombination vary by several orders of magnitude across the genome, and are known to differ between species and even between populations. Studying the differences in recombination maps across populations has been stymied by the confounding effect of differences in demographic history. To address this problem, we developed a method that infers fine-scale recombination rates while taking demography into account and applied our method to infer population-specific recombination maps for each of 26 diverse human populations. These maps recapitulate many aspects of the history of these populations including signatures of the trans-Atlantic slave trade and the Iberian colonization of the Americas. We also investigated modulators of the local recombination rate, finding an unexpected role for Polycomb-group proteins and the tri-methylation of H3K27 in elevating recombination rates. Further differences in the recombination landscape across the genome and between populations are driven by variation in the gene that encodes the DNA-binding protein PRDM9, and we quantify the weak effect of meiotic drive acting to remove its binding sites.

evolutionary biology↗

bioRxiv2018

The key parameters that govern translation efficiency

Translation of mRNA into protein is a fundamental yet complex biological process with multiple factors that can potentially affect its efficiency. In particular, different genes can have quite different initiation rates, while site-specific elongation rates can vary substantially along a given transcript. Here, we analyze a stochastic model of translation dynamics to identify the key parameters that govern the overall rate of protein synthesis and the efficiency of ribosome usage. The mathematical model we study is an interacting particle system that generalizes the Totally Asymmetric Simple Exclusion Process (TASEP), where particles correspond to ribosomes. While the TASEP and its variants have been studied for the past several decades through simulations and mean field approximations, a general analytic solution has remained challenging to obtain. By analyzing the so-called hydrodynamic limit, we here obtain exact closed-form expressions for stationary currents and particle densities that agree well with Monte Carlo simulations. In addition, we provide a complete characterization of phase transitions in the system. Surprisingly, phase boundaries depend on only four parameters: the particle size, and the first, last and minimum particle jump rates. Relating these theoretical results to translation, we formulate four design principles that detail how to tune these parameters to optimize translation efficiency in terms of protein production rate and resource usage. We then analyze ribosome profiling data of S. cerevisiae and demonstrate that its translation system is generally efficient, consistent with the design principles we found. We discuss implications of our findings on evolutionary constraints and codon usage bias.

genomics↗

bioRxiv2018

Differences in the path to exit the ribosome across the three domains of life

Recent advances in biological imaging have led to a surge of fine-resolution structures of the ribosome from diverse organisms. Comparing these structures, especially the exit tunnel, to characterize the key similarities and differences across species is essential for various important applications, such as designing antibiotic drugs and understanding the intricate details of translation dynamics. Here, we compile and compare 20 fine-resolution cryo-EM and X-ray crystallography structures of the ribosome recently obtained from all three domains of life (bacteria, archaea and eukarya). We first show that a hierarchical clustering of tunnel shapes closely reflects the species phylogeny. Then, by analyzing the ribosomal RNAs and proteins localized near the tunnel, we explain the observed geometric variations and show direct association between the conservations of the geometry, structure, and sequence. We find that the tunnel is more conserved in its upper part, from the polypeptide transferase center to the constriction site. In the lower part, tunnels are significantly narrower in eukaryotes than in bacteria, and we provide evidence for the existence of a second constriction site in eukaryotic tunnels. We also show that ribosomal RNA and protein sequences are more likely to be conserved closer to the tunnel, as is the presence of positively charged amino acids. Overall, our comparative analysis shows how the geometric and biophysical properties of the exit tunnel play an important role in ensuring proper transit of the nascent polypeptide chain, and may explain the differences observed in several co-translational processes across species.

biophysics↗

bioRxiv2018

Efficiently inferring the demographic history of many populations with allele count data

The sample frequency spectrum (SFS), or histogram of allele counts, is an important summary statistic in evolutionary biology, and is often used to infer the history of population size changes, migrations, and other demographic events affecting a set of populations. The expected multipopulation SFS under a given demographic model can be efficiently computed when the populations in the model are related by a tree, scaling to hundreds of populations. Admixture, back-migration, and introgression are common natural processes that violate the assumption of a tree-like population history, however, and until now the expected SFS could be computed for only a handful of populations when the demographic history is not a tree. In this article, we present a new method for efficiently computing the expected SFS and linear functionals of it, for demographies described by general directed acyclic graphs. This method can scale to more populations than previously possible for complex demographic histories including admixture. We apply our method to an 8-population SFS to estimate the timing and strength of a proposed \"basal Eurasian\" admixture event in human history. We implement and release our method in a new open-source software package momi2.

evolutionary biology↗

bioRxiv2018

Evaluating Tumor Evolution via Genomic Profiling of Individual Tumor Spheroids in a Malignant Ascites from a Patient with Ovarian Cancer Using a Laser-aided Cell Isolation Technique

BackgroundEpithelial ovarian cancer (EOC) is a silent but mostly lethal gynecologic malignancy. Most patients present with malignant ascites and peritoneal seeding at diagnosis. In the present study, we used a laser-aided isolation technique to investigate the clonal relationship between the primary tumor and tumor spheroids found in the malignant ascites of an EOC patient. Somatic alteration profiles of ovarian cancer-related genes were determined for eight spatially separated samples from primary ovarian tumor tissues and ten tumor spheroids from the malignant ascites using next-generation sequencing.\n\nResultsWe observed high levels of intra-tumor heterogeneity (ITH) in copy number alterations (CNAs) and single-nucleotide variants (SNVs) in the primary tumor and the tumor spheroids. As a result, we discovered that tumor cells in the primary tissues and the ascites were genetically different lineages. We categorized the CNAs and SNVs into clonal and subclonal alterations according to their distribution among the samples. Also, we identified focal amplifications and deletions in the analyzed samples. For SNVs, a total of 171 somatic mutations were observed, among which 66 were clonal mutations present in both the primary tumor and the ascites, and 61 and 44 of the SNVs were subclonal mutations present in only the primary tumor or the ascites, respectively.\n\nConclusionsBased on the somatic alteration profiles, we constructed phylogenetic trees and inferred the evolutionary history of tumor cells in the patient. The phylogenetic trees constructed using the CNAs and SNVs showed that two branches of the tumor cells diverged early from an ancestral tumor clone during an early metastasis step in the peritoneal cavity. Our data support the monophyletic spread of tumor spheroids in malignant ascites.

cancer biology↗

bioRxiv2018

High-throughput inference of pairwise coalescence times identifies signals of selection and enriched disease heritability

Interest in reconstructing demographic histories has motivated the development of methods to estimate locus-specific pairwise coalescence times from whole-genome sequence data. We developed a new method, ASMC, that can estimate coalescence times using only SNP array data, and is 2-4 orders of magnitude faster than previous methods when sequencing data are available. We were thus able to apply ASMC to 113,851 phased British samples from the UK Biobank, aiming to detect recent positive selection by identifying loci with unusually high density of very recent coalescence times. We detected 12 genome-wide significant signals, including 6 loci with previous evidence of positive selection and 6 novel loci, consistent with coalescent simulations showing that our approach is well-powered to detect recent positive selection. We also applied ASMC to sequencing data from 498 Dutch individuals (Genome of the Netherlands data set) to detect background selection at deeper time scales. We observed highly significant correlations between average coalescence time inferred by ASMC and other measures of background selection. We investigated whether this signal translated into an enrichment in disease and complex trait heritability by analyzing summary association statistics from 20 independent diseases and complex traits (average N=86k) using stratified LD score regression. Our background selection annotation based on average coalescence time was strongly enriched for heritability (p = 7x10-153) in a joint analysis conditioned on a broad set of functional annotations (including other background selection annotations), meta-analyzed across traits; SNPs in the top 20% of our annotation were 3.8x enriched for heritability compared to the bottom 20%. These results underscore the widespread effects of background selection on disease and complex trait heritability.

genetics↗

bioRxiv2018

A Likelihood-Free Inference Framework for Population Genetic Data using Exchangeable Neural Networks

An explosion of high-throughput DNA sequencing in the past decade has led to a surge of interest in population-scale inference with whole-genome data. Recent work in population genetics has centered on designing inference methods for relatively simple model classes, and few scalable general-purpose inference techniques exist for more realistic, complex models. To achieve this, two inferential challenges need to be addressed: (1) population data are exchangeable, calling for methods that efficiently exploit the symmetries of the data, and (2) computing likelihoods is intractable as it requires integrating over a set of correlated, extremely high-dimensional latent variables. These challenges are traditionally tackled by likelihood-free methods that use scientific simulators to generate datasets and reduce them to hand-designed, permutation-invariant summary statistics, often leading to inaccurate inference. In this work, we develop an exchangeable neural network that performs summary statistic-free, likelihood-free inference. Our frame-work can be applied in a black-box fashion across a variety of simulation-based tasks, both within and outside biology. We demonstrate the power of our approach on the recombination hotspot testing problem, outperforming the state-of-the-art.

evolutionary biology↗

bioRxiv2017

Geometry of the sample frequency spectrum and the perils of demographic inference

The sample frequency spectrum (SFS), which describes the distribution of mutant alleles in a sample of DNA sequences, is a widely used summary statistic in population genetics. The expected SFS has a strong dependence on the historical population demography and this property is exploited by popular statistical methods to infer complex demographic histories from DNA sequence data. Most, if not all, of these inference methods exhibit pathological behavior, however. Specifically, they often display runaway behavior in optimization, where the inferred population sizes and epoch durations can degenerate to 0 or diverge to infinity, and show undesirable sensitivity of the inferred demography to perturbations in the data. The goal of this paper is to provide theoretical insights into why such problems arise. To this end, we characterize the geometry of the expected SFS for piecewise-constant demographic histories and use our results to show that the aforementioned pathological behavior of popular inference methods is intrinsic to the geometry of the expected SFS. We provide explicit descriptions and visualizations for a toy model with sample size 4, and generalize our intuition to arbitrary sample sizes n using tools from convex and algebraic geometry. We also develop a universal characterization result which shows that the expected SFS of a sample of size n under an arbitrary population history can be recapitulated by a piecewise-constant demography with only{kappa} n epochs, where{kappa} n is between n/2 and 2n - 1. The set of expected SFS for piecewise-constant demographies with fewer than{kappa} n epochs is open and non-convex, which causes the above phenomena for inference from data.

evolutionary biology↗