Bioinformatics

Sloan-Kettering Inst Can Research, New York NY

Investigators

Linked publications, trials & patents

Abstract

The goal of the Bioinformatics core is to provide expert computational analysis of molecular[unreadable] profiling (expression and NMR) data in order to determine the molecular signatures predictive of diagnosis[unreadable] and outcome in Soft Tissue Sarcoma (STS). The core will not only provide computational/statistical analysis[unreadable] but will build and maintain the data infrastructure needed by the various projects, whose work will lead to the[unreadable] definition of new marker sets, mechanistic hypotheses and possible identification of new drug targets. The[unreadable] core will also facilitate integration of research in the projects by enabling the sharing of the various datasets[unreadable] collected. Specifically, it will perform the following tasks. 1) Statistical analysis of microarray expression data[unreadable] including: error analysis, normalization, unsupervised clustering analysis, differential gene analysis and[unreadable] multivariate class prediction. These methods will be applied in the following cases: a. Cluster and differential[unreadable] gene expression analysis of sarcoma subtypes to classify sarcoma tissue samples based on their similarity[unreadable] in gene expression, to identify potential diagnostic/prognostic markers and to determine the relevant genes[unreadable] for subsequent pathway analysis; b. Expression analysis of SYT-SSX regulated genes along with the[unreadable] analysis of the respective promoters and expression based survival prediction of Synovial Sarcomas; c.[unreadable] Supervised learning analysis of clinical variables such as distant recurrence and survival, the object being to[unreadable] generate expression based predictors. 2) Statistical analysis of NMR data obtained from Liposarcoma[unreadable] samples, including prediction of Liposarcoma subtypes and sample clinical variables (outcome/survival)[unreadable] using supervised machine learning techniques. Development of integrated (microarray/NMR) molecular[unreadable] profiling analysis to develop prognostic marker sets. 3) Pathway analysis of molecular profiling data.[unreadable] Integrating data from (1) and (2) with pathway data to: a. Elucidate the biological basis of tumor subtypes; b.[unreadable] Find new potential drug targets. 4) To develop an online repository of microarray expression data along with[unreadable] a database of annotation information and clinical data. Integrate and make available the large collection of[unreadable] datasets to be collected. 5) To develop a patient data tracking system for multi-institutional clinical trials.