Biology subjects

Zong, L.

Publications and source records attributed to Zong, L..

8 recordsLinked to original sources

bioRxiv2026

Biological meaning in protein embedding space is resolution-dependent

Protein language model embeddings are increasingly used to organise biological sequences, yet how biological meaning is encoded within embedding neighbourhoods remains poorly understood. Using two independent hierarchical enzyme systems, carbohydrate-active enzymes and peptidases, we investigated how biological interpretation changes across embedding organisations aligned to different levels of biological hierarchy. Different embedding organisations give rise to distinct neighbourhood semantics. When aligned to membership-boundary resolution, embeddings robustly separated artefacts and unrelated proteins from members of the target category. However, embeddings aligned to functional-grouping resolution maintained compositional neighbourhood structure for multi-domain proteins spanning more than one functional or catalytic group. Finally, embeddings aligned to local-family resolution recovered compact family-like neighbourhoods, including families withheld from training, while weakening broader membership-boundary and functional-grouping relationships. Moreover, embeddings optimised toward the same level of biological organisation retain different biological relationships depending on optimisation trajectory employed. Together, our results show that proximity in protein embedding space has no fixed biological interpretation. Instead, biological meaning emerges across embedding resolutions through selective preservation of different forms of biological organisation.

bioinformatics↗

bioRxiv2026

Popformer: Learning general signatures of positive selection with a self-supervised transformer

Understanding natural selection can help shed light on the genetics underpinning adaptive evolution. The widespread availability of large-scale human genetic variation data has led to the development of data-driven methods for detecting signatures of selection, many of which are based on deep learning. However, these methods often fail to generalize well to the diversity of selection signatures across a broad range of evolutionary scenarios. We propose a novel transformer-based model, Popformer, for learning encodings of general patterns of genetic variation. Popformer includes site-wise and haplotype-wise attention, allowing us to capture variation among both genetic positions and individuals. It additionally learns relative positional embeddings for inter-SNP distances. The model is pre-trained with an analog of the masked language modeling objective across a range of real human genomic data, similar to the task of genetic imputation. Using dimensionality reduction, we show that the pre-trained model learns meaningful embeddings of genomic windows that correspond with population structure. We also show that the model can accurately perform genotype imputation. We further fine-tune the model on selection classification and demonstrate that our model is more accurate than other selection classification methods, on selection simulations of both well-specified and mis-specified demographic models. Using a novel real data validation approach, we apply Popformer to human data from the 1000 Genomes Project and reveal its ability to generalize from simulated data to the diversity of real data. Overall, our method provides a new direction for population genetic inference methods, with future fine-tuning applications including the inference of recombination rates, introgression, and local ancestry. Author summaryNatural selection lies at the heart of adaptation and evolution. Identifying genomic signals of natural selection can lead to insights about evolutionary pressures acting on populations, and provide an understanding of how populations differentiate. Detecting these signals using only present-day genomic data is difficult, and methods developed for this task often use specialized simulations, trading generalization for power. We show here that adapting deep learning techniques from modern language and image processing can be used to identify signals of selection with both high power and robustness to misspecification. By pre-training our transformer model, Popformer, on real data, we induce learned model representations that are representative of the variation in real data. We fine-tune our model on simulations inferred from European populations, and compare with other machine learning methods and summary statistics. Our results suggest that both our model architecture and our training regime are useful for strong performance in diverse simulation tests and in recovering known selection signatures in real data.

genomics↗

bioRxiv2025

Interpreting Convolutional Neural Networks in Population Genetics

AO_SCPLOWBSTRACTC_SCPLOWMachine learning approaches have become a powerful alternative to traditional methods in population genetics. Convolutional neural networks (CNNs) in particular have been successful in inferring natural selection, recombination rate estimation, introgression, dispersal distances, and effective population size changes. One limitation of CNNs and other deep learning methods is that they can be difficult to interpret. When they have been shown to be as or more successful than summary-statistic-based methods, what are they learning? Here we investigate CNNs from two different methods: the pg-gan discriminator for identifying real vs. simulated data, and two networks trained to detect selective sweeps. We first compute correlations between learned network features and traditional summary statistics, then assess whether summary statistics can be predicted from the learned features. To understand the learned features, we compute feature importance through SHAP values and feature groupings through dimensionality reduction. Finally we use decision trees and random forests to build an interpretable "model-of-the-model". Our results reveal that some CNN architectures can implicitly compute summary statistics such as pairwise heterozygosity, while statistics such as the site frequency spectrum are less similar to the networks learned features. We find that long-range linkage disequilibrium is readily approximated by the networks and may be more efficiently computed by CNNs than traditional methods (which are quadratic in the number of sites). Overall, this work contributes to the interpretability of deep learning methods in population genetics by clarifying the relationships between model architecture, learned network features, established summary statistics, and predicted evolutionary parameters.

genomics↗

bioRxiv2024

Semaphorin 4A maintains functional diversity of the hematopoietic stem cell pool

Somatic stem cell pools are comprised of diverse, highly specialized subsets whose individual contribution is critical for the overall regenerative function. In the bone marrow, myeloid-biased HSC (myHSC) are indispensable for replenishment of myeloid cells and platelets during inflammatory response but at the same time, become irreversibly damaged during inflammation and aging. Here, we identify an extrinsic factor, Semaphorin 4A (Sema4A), which non cell-autonomously confers myHSC resilience to inflammatory stress. We show that the absence of Sema4A, myHSC inflammatory hyper-responsiveness in young mice drives excessive myHSC expansion, myeloid bias and profound loss of regenerative function with age. Mechanistically, Sema4A is mainly produced by neutrophils, signals via a cell surface receptor Plexin D1 and safeguards myHSC epigenetic state. Our study shows that by selectively protecting a distinct stem cell subset, an extrinsic factor preserves functional diversity of somatic stem cell pool throughout organismal lifespan.

cell biology↗

bioRxiv2022

Interpretable RNA Foundation Model from Unannotated Data for Highly Accurate RNA Structure and Function Predictions

Non-coding RNA structure and function are essential to understanding various biological processes, such as cell signaling, gene expression, and post-transcriptional regulations. These are all among the core problems in the RNA field. With the rapid growth of sequencing technology, we have accumulated a massive amount of unannotated RNA sequences. On the other hand, expensive experimental observatory results in only limited numbers of annotated data and 3D structures. Hence, it is still challenging to design computational methods for predicting their structures and functions. The lack of annotated data and systematic study causes inferior performance. To resolve the issue, we propose a novel RNA foundation model (RNA-FM) to take advantage of all the 23 million non-coding RNA sequences through self-supervised learning. Within this approach, we discover that the pre-trained RNA-FM could infer sequential and evolutionary information of non-coding RNAs without using any labels. Furthermore, we demonstrate RNA-FMs effectiveness by applying it to the downstream secondary/3D structure prediction, SARS-CoV-2 genome structure and evolution prediction, protein-RNA binding preference modeling, and gene expression regulation modeling. The comprehensive experiments show that the proposed method improves the RNA structural and functional modelling results significantly and consistently. Despite only being trained with unlabelled data, RNA-FM can serve as the foundational model for the field.

bioinformatics↗

bioRxiv2022

DeepAcr: Predicting Anti-CRISPR with Deep Learning

As an important group of proteins discovered in phages, anti-CRISPR inhibits the activity of the immune system of bacteria (i.e., CRISPR-Cas), showing great potential for gene editing and phage therapy. However, the prediction and discovery of anti-CRISPR are challenging for its high variability and fast evolution. Existing biological studies often depend on known CRISPR and anti-CRISPR pairs, which may not be practical considering the huge number of pairs in reality. Computational methods usually struggle with prediction performance. To tackle these issues, we propose a novel deep neural network for anti-CRISPR analysis (AcrNET), which achieves impressive performance. On both the cross-fold and cross-dataset validation, our method outperforms the previous state-of-the-art methods significantly. Impressively, AcrNET improves the prediction performance by at least 15% regarding the F1 score for the cross-dataset test. Moreover, AcrNET is the first computational method to predict the detailed anti-CRISPR classes, which may help illustrate the anti-CRISPR mechanism. Taking advantage of a Transformer protein language model pre-trained on 250 million protein sequences, AcrNET overcomes the data scarcity problem. Extensive experiments and analysis suggest that Transformer model feature, evolutionary feature, and local structure feature complement each other, which indicates the critical properties of anti-CRISPR proteins. Combined with AlphaFold prediction, further motif analysis and docking experiments demonstrate that AcrNET captures the evolutionarily conserved pattern and the interaction between anti-CRISPR and the target implicitly. With the impressive prediction capability, AcrNET can serve as a valuable tool for anti-CRISPR study and new anti-CRISPR discovery, with a free webserver at https://proj.cse.cuhk.edu.hk/aihlab/AcrNET/.

bioinformatics↗

bioRxiv2021

HMD-AMP: Protein Language-Powered Hierarchical Multi-label Deep Forest for Annotating Antimicrobial Peptides

Identifying the targets of an antimicrobial peptide is a fundamental step in studying the innate immune response and combating antibiotic resistance, and more broadly, precision medicine and public health. There have been extensive studies on the statistical and computational approaches to identify (i) whether a peptide is an antimicrobial peptide (AMP) or a non-AMP and (ii) which targets are these sequences effective to (Gram-positive, Gram-negative, etc.). Despite the existing deep learning methods on this problem, most of them are unable to handle the small AMP classes (anti-insect, anti-parasite, etc.). And more importantly, some AMPs can have multiple targets, which the previous methods fail to consider. In this study, we build a diverse and comprehensive multi-label protein sequence database by collecting and cleaning amino acids from various AMP databases. To generate efficient representations and features for the small classes dataset, we take advantage of a protein language model trained on 250 million protein sequences. Based on that, we develop an end-to-end hierarchical multi-label deep forest framework, HMD-AMP, to annotate AMP comprehensively. After identifying an AMP, it will further predict what targets the AMP can effectively kill from eleven available classes. Extensive experiments suggest that our framework outperforms state-of-the-art models in both the binary classification task and the multi-label classification task, especially on the minor classes. Compared with the previous deep learning methods, our method improves the performance on macro-AUROC by 11%. The model is robust against reduced features and small perturbations and produces promising results. We believe HMD-AMP will both contribute to the future wet-lab investigations of the innate structural properties of different antimicrobial peptides and build promising empirical underpinnings for precise medicine with antibiotics.

bioinformatics↗

bioRxiv2021

Self-supervised contrastive learning for integrative single cell RNA-seq data analysis

Single-cell RNA-sequencing (scRNA-seq) has become a powerful tool to reveal the complex biological diversity and heterogeneity among cell populations. However, the technical noise and bias of the technology still have negative impacts on the downstream analysis. Here, we present a self-supervised Contrastive LEArning framework for scRNA-seq (CLEAR) profile representation and the downstream analysis. CLEAR overcomes the heterogeneity of the experimental data with a specifically designed representation learning task and thus can handle batch effects and dropout events. In the task, the deep learning model learns to pull together the representations of similar cells while pushing apart distinct cells, without manual labeling. It achieves superior performance on a broad range of fundamental tasks, including clustering, visualization, dropout correction, batch effect removal, and pseudo-time inference. The proposed method successfully identifies and illustrates inflammatory-related mechanisms in a COVID-19 disease study with 43,695 single cells from peripheral blood mononuclear cells. Further experiments to process a million-scale single-cell dataset demonstrate the scalability of CLEAR. This scalable method generates effective scRNA-seq data representation while eliminating technical noise, and it will serve as a general computational framework for single-cell data analysis.

bioinformatics↗