Biology subjects

Yang, K. K.

Publications and source records attributed to Yang, K. K..

26 records · Page 2Linked to original sources

bioRxiv2022

Deep self-supervised learning for biosynthetic gene cluster detection and product classification

Natural products are chemical compounds that form the basis of many therapeutics used in the pharmaceutical industry. In microbes, natural products are synthesized by groups of colocalized genes called biosynthetic gene clusters (BGCs). With advances in high-throughput sequencing, there has been an increase of complete microbial isolate genomes and metagenomes, from which a vast number of BGCs are undiscovered. Here, we introduce a self-supervised learning approach designed to identify and characterize BGCs from such data. To do this, we represent BGCs as chains of functional protein domains and train a masked language model on these domains. We assess the ability of our approach to detect BGCs and characterize BGC properties in bacterial genomes. We also demonstrate that our model can learn meaningful representations of BGCs and their constituent domains, detect BGCs in microbial genomes, and predict BGC product classes. These results highlight self-supervised neural networks as a promising framework for improving BGC prediction and classification. Author summaryBiosynthetic gene clusters (BGCs) encode for natural products of diverse chemical structures and function, but they are often difficult to discover and characterize. Many bioinformatic and deep learning approaches have leveraged the abundance of genomic data to recognize BGCs in bacterial genomes. However, the characterization of BGC properties remains the main bottleneck in identifying novel BGCs and their natural products. In this paper, we present a self-supervised masked language model that learns meaningful representations of BGCs with improved downstream detection and classification.

bioinformatics↗

bioRxiv2022

Masked inverse folding with sequence transfer for protein representation learning

AO_SCPLOWBSTRACTC_SCPLOWSelf-supervised pretraining on protein sequences has led to state-of-the art performance on protein function and fitness prediction. However, sequence-only methods ignore the rich information contained in experimental and predicted protein structures. Meanwhile, inverse folding methods reconstruct a proteins amino-acid sequence given its structure, but do not take advantage of sequences that do not have known structures. In this study, we train a masked inverse folding protein masked language model parameterized as a structured graph neural network. During pretraining, this model learns to reconstruct corrupted sequences conditioned on the backbone structure. We then show that using the outputs from a pretrained sequence-only protein masked language model as input to the inverse folding model further improves pretraining perplexity. We evaluate both of these models on downstream protein engineering tasks and analyze the effect of using information from experimental or predicted structures on performance.

bioengineering↗

bioRxiv2022

Convolutions are competitive with transformers for protein sequence pretraining

Pretrained protein sequence language models have been shown to improve the performance of many prediction tasks, and are now routinely integrated into bioinformatics tools. However, these models largely rely on the Transformer architecture, which scales quadratically with sequence length in both run-time and memory. Therefore, state-of-the-art models have limitations on sequence length. To address this limitation, we investigated if convolutional neural network (CNN) architectures, which scale linearly with sequence length, could be as effective as transformers in protein language models. With masked language model pretraining, CNNs are competitive to and occasionally superior to Transformers across downstream applications while maintaining strong performance on sequences longer than those allowed in the current state-of-the-art Transformer models. Our work suggests that computational efficiency can be improved without sacrificing performance simply by using a CNN architecture instead of a Transformer, and emphasizes the importance of disentangling pretraining task and model architecture.

bioengineering↗

bioRxiv2022

Randomized gates eliminate bias in sort-seq assays

Sort-seq assays are a staple of the biological engineering toolkit, allowing researchers to profile many groups of cells based on any characteristic that can be tied to fluorescence. However, current approaches, which segregate cells into bins deterministically based on their measured fluorescence, introduce systematic bias. We describe a surprising result: one can obtain unbiased estimates by incorporating randomness into sorting. We validate this approach in simulation and experimentally, and describe extensions for both estimating group level variances and for using multi-bin sorters.

bioengineering↗

bioRxiv2021

FLIP: Benchmark tasks in fitness landscape inference for proteins

Machine learning could enable an unprecedented level of control in protein engineering for therapeutic and industrial applications. Critical to its use in designing proteins with desired properties, machine learning models must capture the protein sequence-function relationship, often termed fitness landscape. Existing bench-marks like CASP or CAFA assess structure and function predictions of proteins, respectively, yet they do not target metrics relevant for protein engineering. In this work, we introduce Fitness Landscape Inference for Proteins (FLIP), a benchmark for function prediction to encourage rapid scoring of representation learning for protein engineering. Our curated tasks, baselines, and metrics probe model generalization in settings relevant for protein engineering, e.g. low-resource and extrapolative. Currently, FLIP encompasses experimental data across adeno-associated virus stability for gene therapy, protein domain B1 stability and immunoglobulin binding, and thermostability from multiple protein families. In order to enable ease of use and future expansion to new tasks, all data are presented in a standard format. FLIP scripts and data are freely accessible at https://benchmark.protein.properties.

bioengineering↗

bioRxiv2021

A Topological Data Analytic Approach for Discovering Biophysical Signatures in Protein Dynamics

Identifying structural differences among proteins can be a non-trivial task. When contrasting ensembles of protein structures obtained from molecular dynamics simulations, biologically-relevant features can be easily overshadowed by spurious fluctuations. Here, we present SINATRA Pro, a computational pipeline designed to robustly identify topological differences between two sets of protein structures. Algorithmically, SINATRA Pro works by first taking in the 3D atomic coordinates for each protein snapshot and summarizing them according to their underlying topology. Statistically significant topological features are then projected back onto a user-selected representative protein structure, thus facilitating the visual identification of biophysical signatures of different protein ensembles. We assess the ability of SINATRA Pro to detect minute conformational changes in five independent protein systems of varying complexities. In all test cases, SINATRA Pro identifies known structural features that have been validated by previous experimental and computational studies, as well as novel features that are also likely to be biologically-relevant according to the literature. These results highlight SINATRA Pro as a promising method for facilitating the non-trivial task of pattern recognition in trajectories resulting from molecular dynamics simulations, with substantially increased resolution. Author SummaryStructural features of proteins often serve as signatures of their biological function and molecular binding activity. Elucidating these structural features is essential for a full understanding of underlying biophysical mechanisms. While there are existing methods aimed at identifying structural differences between protein variants, such methods do not have the capability to jointly infer both geometric and dynamic changes, simultaneously. In this paper, we propose SINATRA Pro, a computational framework for extracting key structural features between two sets of proteins. SINATRA Pro robustly outperforms standard techniques in pinpointing the physical locations of both static and dynamic signatures across various types of protein ensembles, and it does so with improved resolution.

biophysics↗

bioRxiv2021

Evolutionary velocity with protein language models

Predicting the order of biological homologs is a fundamental task in evolutionary biology. For protein evolution, this order is often determined by first arranging sequences into a phylogenetic tree, which has limiting assumptions and can suffer from substantial ambiguity. Here, we demonstrate how machine learning algorithms called language models can learn mutational likelihoods that predict the directionality of evolution, thereby enabling phylogenetic analysis that addresses key limitations of existing methods. Our main conceptual advance is to construct a "vector field" of protein evolution through local evolutionary predictions that we refer to as evolutionary velocity (evo-velocity). We show that evo-velocity can successfully predict evolutionary order at vastly different timescales, from viral proteins evolving over years to eukaryotic proteins evolving over geologic eons. Evo-velocity also yields new evolutionary insights, predicting strategies of viral-host immune escape, resolving conflicting theories on the evolution of serpins, and revealing a key role of horizontal gene transfer in the evolution of eukaryotic glycolysis. In doing so, our work suggests that language models can learn sufficient rules of natural protein evolution to enable evolutionary predictability.

bioinformatics↗

bioRxiv2019

Machine learning-guided channelrhodopsin engineering enables minimally-invasive optogenetics

We have engineered light-gated channelrhodopsins (ChRs) whose current strength and light sensitivity enable minimally-invasive neuronal circuit interrogation. Current ChR tools applied to the mammalian brain require intracranial surgery for transgene delivery and implantation of invasive fiber-optic cables to produce light-dependent activation of a small volume of brain tissue [~1 mm3]. To enable optogenetics for large brain volumes and without the need for invasive implants, our ChR engineering approach leverages the significant literature of ChR variants to train statistical models for the design of new, high-performance ChRs. With Gaussian Process models trained on a limited experimental set of 102 functionally characterized ChR variants, we designed high-photocurrent ChRs with unprecedented light sensitivity; three of these, ChRger1, ChRger2, and ChRger3, enable optogenetic activation of the nervous system via minimally-invasive systemic transgene delivery with rAAV-PHP.eB, which was not possible previously due to low per-cell transgene copy produced by systemic delivery. These engineered ChRs enable light-induced neuronal excitation without invasive intracranial surgery for virus delivery or fiber optic implantation, i.e. they enable minimally-invasive optogenetics.

bioengineering↗