Biology subjects

LI, B.

Publications and source records attributed to LI, B..

6 recordsLinked to original sources

bioRxiv2026

Genetic introgression and transcriptomic plasticity are associated with enhanced Leishmania infantum pathogenicity causing human cutaneous leishmaniasis in Tunisia

The protozoan parasite Leishmania infantum exhibits significant genetic variability among isolates, influencing disease manifestation and treatment response. Although L. infantum is classically described as the causative agent of Visceral Leishmaniasis (VL) - often associated with immune deficiency, cases of Cutaneous Leishmaniasis (CL) caused by this species in immunocompetent individuals have been reported in different countries. To investigate the molecular basis of this unusual shift in tissue tropism and pathogenicity, we applied comparative genomic and transcriptomic approaches on two canine isolates (CanL) and two human isolates associated with Cutaneous Leishmaniasis (CL) in Tunisia. While the CanL isolates showed close genetic similarity to the L. infantum reference strain (JPCM5), the CL isolates formed a separate, highly divergent cluster based on SNP localization and frequency, differing not only from JPCM5 but also from each other. Utilizing the metagenomics sequence classification tool Kraken, we revealed a complex hybrid nature of the CL isolates, showing introgression from L. donovani and L. tropica, suggesting that hybridization has played a key role in generating novel phenotypic traits. Integration of RNA-seq and DNA-seq data demonstrated that only a minority of gene expression variation within and in-between the CanL or CL groups reflected gene dosage effects due to copy number variation, while the majority of expression differences were independent of gene dosage, implying post-transcriptional regulatory mechanisms contributing to parasite adaptation. In conclusion, our study identifies hybridization, genome instability, and transcriptomic adaptation as interconnected drivers of the L. infantum evolutionary potential. These mechanisms can collectively enhance parasite fitness gain, potentially explaining the emergence of cutaneous disease forms in a species traditionally linked to visceral infection. Author SummaryThis research reveals that hybridization between distinct Leishmania parasite species could be a key mechanism driving the evolution of new disease forms. By demonstrating that cutaneous leishmaniasis (CL) cases are caused by hybrid L. infantum parasites whose genomes show introgression with DNA from L. donovani and L. tropica, this study reveals a molecular mechanism potentially linked to the emergence of tegumentary disease from a species traditionally known to cause visceral infection. These findings contribute to our understanding of Leishmania evolution, the emergence of atypical forms of leishmaniasis linked to hybridization, and the impact of genome instability and transcriptomic adaptation as potent forces for generating phenotypic diversity and enhancing parasite fitness.

genomics↗

bioRxiv2026

A Multi-modal LLM-Knowledge Fusion Framework for Predicting Single-cell Genetic Perturbation Effects

Understanding cellular responses to genetic perturbations is fundamental for drug discovery, yet experimental approaches face significant limitations in coverage and cost that prevent comprehensive mapping of cellular behavior. This has motivated the development of virtual cells--computational models that learn the relationship between cell state and function to predict the consequences of perturbations across diverse contexts. However, current computational methods suffer from limited accuracy in complex genetic interactions, poor biological interpretability, and inadequate generalization to unseen genes, severely constraining virtual cell capabilities. We present scPert, a multi-modal framework based on Transformer architecture that integrates large language model embeddings with structured biological knowledge to predict single-cell transcriptomic responses to genetic perturbations. Through hierarchical fusion of knowledge graph representations, contextual embeddings from foundation models, and gene-specific encodings, scPert achieves significant performance improvements in both single-gene and combinatorial perturbations over existing methods. In cancer-relevant applications, scPert demonstrates the capability to reveal p53 pathway dynamics and immune checkpoint regulatory mechanisms. Systematic evaluation on 42 cancer dependency genes demonstrates scPerts ability to identify critical potential therapeutic targets. Our framework establishes a powerful computational foundation for virtual cell construction and accelerates drug target discovery.

bioinformatics↗

bioRxiv2024

Discovery and Engineering of a New BvCas12a Nuclease for Mammalian Genome Editing and Nucleic Acid Detection

Cas12a is an RNA-guided endonuclease that has emerged as a powerful gene-editing tool. We have identified a novel Cas protein, BvCas12a, from Butyricimonas virosa with a 5-TYTN protospacer adjacent motif (PAM). BvCas12a exhibits double-stranded DNA cleavage activity in vitro and genome editing activity in eukaryotic cells. Though the editing efficiency of BvCas12a is marginally lower than that of AsCas12a, the editing specificity of BvCas12a in eukaryotic cells is comparable or superior to that of AsCas12a. Moreover, BvCas12a exhibits substantial collateral activity and can detect HPV DNA effectively and accurately in conjunction with isothermal amplification, highlighting its potential in nucleic acid diagnostics. Furthermore, we have engineered BvCas12a to create two variants, BvCas12a-R (N549R, T606P) and BvCas12a-RVR (N549R, K555V, C559R, T606P). These variants recognize expanded 5-YYN and 5-YN PAM in vitro, respectively. Additionally, they exhibit higher editing activity than wild-type BvCas12a and recognize 5-YYN PAM in vivo at all sites detected. In conclusion, we have identified a novel BvCas12a protein with high specificity and engineered two variants with broader PAM compatibility and improved genome editing efficiency. These findings offer a potent gene editing tool for application in scientific research, gene therapy, and nucleic acid diagnostics.

genomics↗

bioRxiv2021

Whole-genome sequencing of Chinese native goat offers biological insights into cashmere fiber formation

Cashmere evolved naturally in the goat, and almost all breeds of goat can produce more or less cashmere fibers. However, the genetic alterations underlying cashmere trait selection are still unclear.We sequenced 120 Chinese native goat including two cashmere goat breeds (Ujumain, Chaidamu) and six ordinary goat breeds (Jining Gray, Matou, Guizhou Black, Jintang Black, Yunnan Black Bone, Chengdu Brown). The genome-wide selective sweep of cashmere goat and ordinary goat revealed a novel set of candidate genes as well as pathways, such as Nuclear factor kappa-B and Wnt Signaling pathways. Of them, the LHX2 gene regulating hair follicle development, was evident from the strongest selection signal when comparing the Uhumqin cashmere goat and ordinary goat. Interestingly, we identified a 582bp deletion at 367 kb upstream of LHX2 with higher frequency in cashmere goats and their ancient relatives. This mutation probably rises along the breeding procedures, and is putatively responsible for cashmere production and diameter, as revealed by association studies. Luciferase assay shows that the deletion, which acts as an insulator, restrains the expression of LHX2 by interfering its upstream enhancers.Our study discovers a novel insulator of the LHX2 involved in regulating cashmere production and diameter, which would be beneficial to understanding hair follicle development and regeneration. Our findings also provide new insights into the genetic formation of cashmere, and facilitate subsequent molecular breeding for cashmere goat improvement.

genomics↗

bioRxiv2021

QSAR Model Based Gradient Boosting Regression of N-Arylsulfonyl-Indole-2-Carboxamide Derivatives as Inhibitors for Fructose-1,6-Bisphosphatase

As is known to all, diabetes metellius is a global health threaten and it has caused worldwide attention of scientists. To get a better investigation of the drug design of diabetes, we used heuristic method to established the linear model and used Gradient Boosting Regression to establish the nonlinear model of Fructose-1,6-Bisphosphatse inhibitor successively. In this study, 84 derivatives of N-Arylsulfonyl-Indole-2-Carboxamide were introduced into the models, two outstanding QSAR models with 2 molecule descriptors were established successfully. Grandient Boosting Regression rendered a good correlation with R2 of 0.943 and MSE of 0.135 for the training set, 0.916 and 0.213 for test set, which also proves the feasibility of the implementation of the new method GBR in the field of QSAR. Meanwhile, the optimal model displayed wonderful statistical significance. This study shows unlimited potential for design of new drugs for diabetes.

biochemistry↗

bioRxiv2020

A multi-task deep-learning system for predicting membrane associations and secondary structures of proteins

Accurate prediction of secondary structures and transmembrane segments is often the first step towards modeling the tertiary structure of a protein. Existing methods are either specialized in one class of proteins or developed to predict one type of 1D structural attributes (secondary structure, topology, or transmembrane segment). In this work, we develop a new method for simultaneous prediction of secondary structure, transmembrane segment, and transmembrane topology with no a priori assumption on the class of the input protein sequence. The new method, Membrane Association and Secondary Structures of Proteins (MASSP) predictor, uses multi-tiered neural networks that incorporate recent innovations in machine learning. The first tier is a multi-task multi-layer convolutional neural network (CNN) that learns patterns in image-like input position-specific-scoring matrices (PSSMs) and predicts residue-level 1D structural attributes. The second tier is a long short-term memory (LSTM) neural network that treats the predictions of the first tier from the perspective of natural language processing and predicts the class of the input protein sequence. We curated a non-redundant data set consisting of 54 bitopic, 241 multi-spanning TM-alpha, 77 TM-beta, and 372 soluble proteins, respectively for training and testing MASSP. For secondary structure prediction, the mean three-state accuracy (Q3) of MASSP is 0.830, better than the Q3 of PSIPRED (0.829) and that of SPINE-X (0.813) and substantially better than that of Jufo9D (0.762) and RaptorX-Property (0.741). The mean segment overlap score (SOV) of MASSP is 0.752, gaining at least 7.7% improvement over all the other four methods. For transmembrane topology prediction, MASSP has a performance comparable to OCTOPUS and substantially better than MEMSAT3 and TMHMM2 on TM-alpha proteins, and on TM-beta proteins, MASSP is significantly better than both BOCTOPUS2 and PRED-TMBB2. By integrating prediction of secondary structure and transmembrane segments in a deep-learning framework, MASSP improves performance over previous methods, has broader applicability, and enables proteome scale predictions.

bioinformatics↗