TIME SERIES METHODS FOR UNDERSTANDING DYNAMICAL BIOLOGICAL SYSTEMS

Abstract

The goal of this seedling is to develop methods for learning and inference from noisy, high-dimensional time series and combining it with prior knowledge and constraints to develop new insights about biological systems, and how they change over time based on specific signals. As a concrete example, consider the task of learning the process by which a skin cell converts into a muscle cell. Although we have a general model for cellular differentiation its details specific to cells are still unknown and important to clarify for understanding human disease and for synthetic biology. The recent technological improvement that breaks these limitations is the development of single-cell RNA-seq (scRNA-seq) and single-cell assay for transposase-accessible chromatin by sequencing (scATAC-seq). With single-cell -omics data, it is accomplishable to computationally order cells along trajectories, allowing the unbiased study of cellular dynamic processes. On one hand, scRNA-seq gives gene expression variation of the whole genome across time in thousands of single cells; on the other hand, scATAC-seq measures chromatin accessibility in thousands of single cells per assay. A very recent work done by Buenrostro et al. has shown that the expression level of a gene given by pseudotime scRNA-seq data actually have the same temporal pattern as its corresponding TF motif accessibility dynamics across myeloid pseudotime of scATAC-seq data. This general model identifies the target genes of a TF by linking the temporal pattern of TF motif accessibility and genes expression level.Learning a gene regulatory network from it will thus lead to a result that gives insights to the transcriptional program in a more causally interpreted way. The regulatory effects on target genes of a transcription factor can also be listed in chronological order, providing information for a more detailed biological process.

Document Details

Document Type
DoD Grant Award
Publication Date
Oct 06, 2018
Source ID
W911NF1710222

Entities

People

  • Suchi Saria

Organizations

  • Army Contracting Command
  • Defense Advanced Research Projects Agency
  • Johns Hopkins University

Tags

Fields of Study

  • Biology

Readers

  • Molecular and genetic basis of cancer.
  • Neural Network Machine Learning.
  • Systems Analysis and Design

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks
  • Biotechnology