Temporal and spatial dynamics in (transcript)omics experimentation

General info

Date from - to

01 Jan 2010 - 01 Dec 2013

Project leader(s)

Breit, Timo Dr.

Abstract

Aim of the project:
This project aims to prove that well-designed omics experiments, based on newly gained knowledge of transcriptome dynamics in time, space and in response to treatment, will allow for deeper insight in the cellular mechanisms at hand.

Key objectives:

• a bioinformatics pipeline to integrate various redundant/conflicting genomics resources for integration on sequence level towards microarray design.
• methods for arranging measurement points of high-resolution sample series in the correct order, in the presence and absence of phenotypic markers.  
• prove that better understanding of the spatio-temporal dynamics of the transcriptome in combination with better design for systems-level omics experimentation will enhance pathway and network reconstruction.
• show that treatment-dose range finding is essential in whole-genome (transcript)omics experimentation.
• assemble guidelines for optimal design for experimentation with respect to time, space and treatment dose.

Approach:
The spatial and temporal dynamics of the transcriptome are largely unknown. This has an important effect on (transcript)omics experimental designs. Frequently, most biological variation becomes confounded between dispersed time points or within heterogeneous tissue/cell populations, which severely hampers valid interpretation. To achieve our project aim and key objectives, we will use data from several dedicated biological transcriptomics experiments of matching projects, such as: a high-resolution time series (200 single Zebrafish embryos, 1 per 12 seconds) and subsequent in-situ hybridizations to dissect the transcriptome complexity into a manageable number of activated genes per cell type at each time point. To be able to interpret these high-resolution experiments, we will develop advanced bioinformatics methods for the analysis of the data. These comprise of methods for; arranging high-resolution sample series with and without clear morphological markers for ordering biological samples; non-linear modeling to establish differential gene expression and transcription starting points in time and space in sample series without replicas; visualization models to integrate transcriptomics, in-situ data and pathway/network knowledge. We achieve proof-of-principle by showing that we will find better biological results i.e. more differentially expressed genes, better-defined pathways/genetic networks, less non-specific responses, and/or more detailed knowledge on involved cellular processes, compared to state-of-the-art studies on the same biological topics.

Publications

• Integrating heterogeneous sequence information for transcriptome-wide microarray design; a Zebrafish example