Time: 16:00-17:30 pm, Mar.20(Wednesday)
Venue: Room 300, SIBS Main Building, 320 Yueyang Road
Host: PICB Student Group
Speaker1: Jinxi Li
Topic 1:Variants in limb development control genes underline systematic diversity variations in human fingerprint patterns
Abstract: The human fingerprint, a complex pattern of arches, loops and whorls, is of longstanding practical and cultural interest, and has become a metaphor for uniqueness across many domains. While of great interest in many aspects of human activity, little is known about the genes that underlie variation in fingerprint patterns and the developmental mechanisms that they influence. Here we report a genome-wide association study in Chinese populations to understand the basis of variations in the fingerprint pattern, finding three independent associated loci (3q26.2, 18q23 and 8q23.1), with nearby genes (EVI1, SALL3 and RSPO2) each linked to limb development. Further, we clarify the influence of genetic variants on fingerprint pattern by using bioinformatics, molecular biology and mice model experiments. Moreover, trans-ethnic meta-analysis of more than 16,000 individuals further revealed 17 novel loci, strongly enriched in genes operating in general limb development pathways. Our findings provide novel insight in the genetics of the fingerprint pattern and provide important clues to its developmental mechanisms in digit growth and shaping.
Topic 2: An accurate machine learning based screening model of Down's syndrome based on distinctive characteristics of dermatoglyphic features
Abstract: Chromosomal disorders are usually associated with distinctive characteristics of dermatoglyphic features. For example, simian lines are often observed in Down’s syndrome patients. However, no quantitative models have been tested in a rigorous study design. Here we report a comprehensive study on the distinctive characteristics of 56 dermatoglyphic features in 256 Down’s syndrome patients and 800 controls. By using eXtreme Gradient Boosting, we build a machine learning based model to screen Down’s syndrome based on 6 most informative dermatoglyphic features. Our model is high in accuracy (98.04%), with a relatively low false negative rate (6.67%). Since the dermatoglyphic features usually remain unchanged through the entire life, our model can help screen Down’s syndrome at the time of birth, providing feasibility for early diagnosis and intervention. This study also sets an example for studying the implication of dermatoglyphic features in other chromosomal disorders.
Speaker2: Ke Zhang
Title: Single-cell graph cuts optimization (scGCO) for identifying spatially variable genes
Abstract: Single-cell RNA sequencing (scRNA-seq) has emerged as a powerful technologies for studying rare cells and delineating complex populations, which recovers cell-to-cell commutation and complex tissues heterogeneity, such as cellular development and pathogene evolution. Single-cell gene expression data with positional information are critical to dissect mechanisms and architectures of multicellular organisms, from which we have not only insight into functions and phenotypes of cells, but also reveal pathological changes and cancer tissue formation in individual cells level. However, limited by current analysis algorithms, we have developed a novel algorithm---scGCO, a method based on fast optimization of Markov Random Fields with graph cuts, to identify spatially viable genes. scGCO delivers superior performance with optimal segmentation of spatial patterns, and can process millions of cells due to its linear sclability.