Biology subjects

Kuang, T.

Publications and source records attributed to Kuang, T..

6 recordsLinked to original sources

bioRxiv2026

Phenotype-driven de novo molecular design from gene expression signatures

AO_SCPLOWBSTRACTC_SCPLOWTarget-based and structure-guided drug design remain central to modern drug discovery, but complementary strategies are needed when predefined targets or binding pockets do not fully capture disease biology. Gene-expression signatures provide scalable system-level readouts of disease and perturbation states, making them attractive inputs for phenotype-guided molecular design. However, preserving phenotypic information during molecular generation remains challenging, and chemically plausible molecules may lose connection to the intended biological response. Here, we present Tx2Mol, a transcriptome-guided framework that translates gene-expression signatures into candidate molecules while maintaining biological guidance throughout generation. We evaluated Tx2Mol across three biological settings: bulk gene perturbation, single-cell perturbation, and patient-derived disease signatures; and three validation dimensions: chemical plausibility, structural compatibility, and phenotypic preservation. Across 10 cancer-relevant bulk gene-perturbation benchmarks, Tx2Mol outperformed 9 transcriptome-guided baselines, improving maximum Tanimoto similarity to known ligands by 24.10% on average and by 50.67% on HDAC1. Structure-based analyses further supported structurally novel candidates with favorable predicted target binding. Tx2Mol also generalized to noisy single-cell perturbation profiles and preserved drug-induced transcriptional responses through in silico drug-perturbation validation. Patient-derived disease signatures further guided molecular generation toward approved-drug chemical space. Together, these results support gene-expression phenotypes as actionable guidance signals for phenotype-directed molecular design and candidate prioritization.

bioinformatics↗

bioRxiv2024

Structures of PSI-FCPI from Thalassiosira pseudonana in high light provide convergent evolution and light-adaptive strategies in diatom FCPIs

Diatoms achieve great survival success in the fluctuating oceanic environment, rely on fucoxanthin chlorophyll a/c-binding proteins (FCPs) to complete light harvesting and quenching, which provide about 20% primary productivity on earth. We report two cryo-electron microscopic structures of photosystem I (PSI) with 13 or 5 FCPIs respectively at 2.78 [A] and 3.20 [A] resolution from Thalassiosira pseudonana under high light conditions. 8 Lhcr FCPIs are found detached from the PSI-13FCPI supercomplex under high light conditions, remaining 5 FCPIs are stably combined with the PSI core including Lhcr3, RedCAP, Lhcq8, Lhcf10, and FCP3 subunits. The specific pigment network in this centric diatom T. pseudonana demonstrates a higher proportion of Chlorophylls a, diadinoxanthins, and diatoxanthins but fewer fucoxanthins compared with the huge PSI-FCPI from another centric diatom Chaetoceros gracilis, thus exhibiting more efficiency in energy transfer and dissipation among FCPI antennas. These results reveal the assembly mechanism of several types of peripheral FCPIs and corresponding light-adaptive strategies in T. pseudonana, as well as the convergent evolution of the diatoms PSI-FCPI structures.

plant biology↗

bioRxiv2023

3D-Mol: A Novel Contrastive Learning Framework for Molecular Property Prediction with 3D Information

AO_SCPLOWBSTRACTC_SCPLOWMolecular property prediction offers an effective and efficient approach for early screening and optimization of drug candidates. Although deep learning based methods have made notable progress, most existing works still do not fully utilize 3D spatial information. This can lead to a single molecular representation representing multiple actual molecules. To address these issues, we propose a novel 3D structure-based molecular modeling method named 3D-Mol. In order to accurately represent complete spatial structure, we design a novel encoder to extract 3D features by deconstructing the molecules into three geometric graphs. In addition, we use 20M unlabeled data to pretrain our model by contrastive learning. We consider conformations with the same topological structure as positive pairs and the opposites as negative pairs, while the weight is determined by the dissimilarity between the conformations. We compare 3D-Mol with various state-of-the-art(SOTA) baselines on 7 benchmarks and demonstrate our outstanding performance in 5 benchmarks.

bioinformatics↗

bioRxiv2023

Pleistocene glaciation advances the younger temporal dimension of species diversification in a major biodiversity hotspot

O_LIThe expansive Himalayan-Hengduan Mountains (HM) and Qinghai-Tibet Plateau (QTP) form a biodiversity hotspot imperiled by global change. The species diversity, resulting from myriad factors led to plant diversification from 10 million years ago till now. However, despite understanding speciation timing better, a lack of in-depth population-level studies and a dearth of organismal sampling among closely related species and populations leaves an incomplete view of diversification trends and biotic influences. C_LIO_LIWe delve into the complex factors influencing diversification through genomic and eco-morphological analysis within the Stellera chamaejasme L. complex, known for its significant floral diversity. C_LIO_LIOur results uncover four cryptic species, indicating a more recent and younger diversification period ([~]2.67-0.9 Mya), driven by Pleistocene glaciation and a complex set of interacting biotic factors. These factors prompted allopatric speciation and advocated cyclical warming-cooling episodes along latitudinal and altitudinal gradients throughout the Pleistocene. C_LIO_LIThe study emphasizes the existence of cryptic species within these mountains, broadening our understanding of species diversification to more recent than previously recognized. This novel perspective may reshape evolutionary paradigms in plant science. Additionally, it also brings to light concerns over future warmings possible impacts on alpine species distribution and speciation, underlining the urgency for increased conservation efforts. C_LI

evolutionary biology↗

bioRxiv2022

Genomic underpinnings of convergent adaptation to high altitudes for alpine plants

Evolutionary convergence is one of the most striking examples of adaptation driven by natural selection. However, genomic evidence for convergent adaptation to extreme environments remains scarce. The Himalaya-Hengduan Mountains represent the worlds most species-rich temperate alpine biota, providing an ideal "natural laboratory" for studying convergent adaptation to high altitudes. Here, we generate reference genomes for two alpine plants, Saussurea obvallata (Asteraceae) and Rheum alexandrae (Polygonaceae), with 37,938 and 61,463 annotated protein-coding genes. By integrating an additional five alpine genomes, we investigate genomic signatures of convergent adaptation to the hostile environments of high altitudes. We show that alpine genomes tend to mitigate their genetic load by contracting genes functioning in the immune system to survive such harsh environments with few pathogens present. We detect signatures of convergent positive selection on a set of genes involved in reproduction and development and reveal that molecular convergence has acted on genes involved in self-incompatibility, cell wall modification, DNA repair and stress resistance, which underlie adaptation to extremely cold, high UV radiation and hypoxia environments. Using gene expression profiles, we further demonstrate that genes associated with cuticular wax and flavonoid biosynthetic pathways exhibit higher expression levels in leafy bracts, shedding lights on the genetic mechanisms of the adaptive greenhouse morphology. Our integrative data provide genomic insights into the convergent evolution at higher-taxonomic levels, aiding in deep understanding of genetic adaptation to complex environments.

evolutionary biology↗

bioRxiv2017

Gene markers for exon capture and phylogenomics in ray-finned fishes

Gene capture coupled with the next generation sequencing has become one of the favorable methods in subsampling genomes for phylogenomic studies. Many target gene markers have been developed in plants, sharks, frogs, reptiles and others, but few have been reported in the ray-finned fishes. Here, we identified a suite of \"single-copy\" protein coding sequence (CDS) markers through comparing eight fish genomes, and tested them empirically in 83 species (33 families and 11 orders) of ray-finned fishes. Sorting through the markers according to their completeness and phylogenetic decisiveness in taxa tested resulted in a selection of 4,434 markers, which were proven to be useful in reconstructing phylogenies of the ray-finned fishes at different taxonomic level. We also proposed a strategy of refining baits (probes) design a posteriori based on empirical data. The markers that we have developed may fill a gap in the tool kit of phylogenomic study in vertebrates.

evolutionary biology↗