Xiao Di, Sahadevan Sonu, Mangala Melissa M, Kim Hani Jieun, Fredericks Anna, Huang Hao, Jothi Raja, Tam Patrick, Gonzalez-Cordero Anai, Zyner Katherine G, Yang Pengyi
Computational Systems Biology Unit, Children's Medical Research Institute, The University of Sydney, Westmead, NSW 2145, Australia.
Stem Cell Medicine Unit, Children's Medical Research Institute, The University of Sydney, Westmead, NSW 2145, Australia.
bioRxiv. 2025 Jul 12:2025.07.09.664003. doi: 10.1101/2025.07.09.664003.
Identifying chemical compounds that target trans-regulatory networks (TRNs) underlying molecular programs of cells for directed cellular conversion (i.e. differentiation, reprogramming, transdifferentiation, and dedifferentiation) is a key step towards advancing regenerative medicine. Recent innovations in single-cell omics technologies enabled high-resolution profiling of TRNs that govern cell identity and cell-fate decisions. Here, we introduce Refate, a computational framework that integrates large-scale multimodal single-cell atlas data to quantify cell propensity of genes, together with six drug databases, to identify chemical compounds that target TRNs for directed cellular conversion. The reconstructed TRNs, including protein-protein interactions and gene regulatory networks, alongside chemical compounds that drive the cellular conversion provide greater biological interpretability and improve efficiency and efficacy. We evaluated Refate by testing its ability to uncover known transcription factors and chemical compounds validated in experimental conversions of various cell types. Furthermore, we experimentally validated the attribute of several novel chemical compounds identified by Refate for enhancing the conversion of human embryonic stem cells to human cranial neural crest cells. Together, these findings demonstrate Refate as an effective tool for discovering chemical compounds that target TRNs to enable cellular conversion, advancing efforts towards regenerative medicine.
识别靶向细胞分子程序背后的转录调控网络（TRNs）以实现定向细胞转化（即分化、重编程、转分化和去分化）的化合物是推进再生医学的关键一步。单细胞组学技术的最新创新使得对控制细胞身份和细胞命运决定的TRNs进行高分辨率分析成为可能。在这里，我们介绍了Refate，这是一个计算框架，它整合大规模多模态单细胞图谱数据以量化基因的细胞倾向，并结合六个药物数据库，来识别靶向TRNs以实现定向细胞转化的化合物。重建的TRNs，包括蛋白质-蛋白质相互作用和基因调控网络，以及驱动细胞转化的化合物，提供了更高的生物学可解释性，并提高了效率和功效。我们通过测试Refate揭示在各种细胞类型的实验转化中得到验证的已知转录因子和化合物的能力来对其进行评估。此外，我们通过实验验证了Refate鉴定出的几种新型化合物促进人类胚胎干细胞向人类颅神经嵴细胞转化的特性。总之，这些发现表明Refate是发现靶向TRNs以实现细胞转化的化合物的有效工具，推动了再生医学的发展。
