Prof. Yu Li’s team developed an AI method to decipher bio-sample heterogeneity, which was published in Nature Communications in Nov 2022. Led by one Ph.D. student, Yixuan Wang, and a RA, Yanshuo Chen, in Prof. Yu Li’s group, this work started in 2021 when they were Year-3 UG students.
Studies have found significant heterogeneity within cancer tumors, containing various cell types. These biological samples are typically deciphered using bulk RNA sequencing (bulk RNA-seq). While bulk RNA-seq can only explain the differences between samples from a holistic perspective, their resolution is insufficient to depict the differences between individual cells. Researchers are, therefore, eager to obtain cell types in the tumor microenvironment at the single-cell level to develop more precise cancer therapies.
In this work, Prof. Yu Li’s team presents an artificial intelligence method, Tissue-AdaPtive autoEncoder (TAPE), a deep learning model that combines bulk RNA-seq with single-cell RNA sequencing (single-cell RNA-seq) to achieve accurate deconvolution in a short time. Compared to existing methods, TAPE can predict the proportions of different cell types in tissues more accurately, and this effectiveness has been mirrored on multiple datasets. Since TAPE is sensitive to cell type changes, for example, it can significantly predict that for COVID-infected tissues, there are fewer pancreatic beta cells than uninfected tissues, and beta cell proportions return to uninfected levels after treating COVID with drugs, they believe that TAPE, developed based on artificial intelligence, can help facilitate the development of precision medicine. Further, TAPE shows its ability to predict cell-type-specific gene expression profiles with biological significance, which helps biologists and medical researchers to perform downstream analysis leading to enlightening biological results and precision medical diagnoses.
The work also demonstrates that artificial intelligence does not just mimic humans but can assist people in solving healthcare problems that humans are temporarily unable to solve.