Clustering and visualization of big and multimodal omics data
Người báo cáo: Đỗ Văn Hoàn (Học viện Kỹ Thuật Quân Sự)

Thời gian: từ 14h, Thứ 6, ngày 3/6/2022

Địa điểm: Phòng 614, Nhà A6

Tóm tắt: Emerging single-cell genomics technologies such as single-cell RNA sequencing provide new opportunities for the discovery of previously unknown cell types and facilitate the study of biological processes such as cancer development. Clustering and visualization using dimensionality reduction techniques such as t-SNE and UMAP are the fundamental steps in analyzing high-dimensional data produced by the technologies. However, computational models have been challenged by the exponential growth of the data thanks to the growth of large-scale genomic projects such as the Human Cell Atlas. In this talk, we will introduce Specter, a computational method that utilizes recent algorithmic advances in fast spectral clustering and ensemble learning. Specter achieves a substantial improvement in accuracy over existing methods and identifies rare cell types with high sensitivity. Moreover, its speed allows Specter to scale to millions of cells and leads to fast computation times in practice. In addition, we will present j-SNE as the generalization to the joint visualization of multimodal omics data, e.g., CITE-seq data that simultaneously measure gene and protein marker expression. The approach automatically learns the relative importance of each modality in order to obtain a concise representation of the data.

Back