Challenges in Computational Functional Genomics Igor Ulitsky Genomics

Challenges in Computational & Functional Genomics Igor Ulitsky

Genomics � “the branch of genetics that studies organisms in terms of their genomes (their full DNA sequences)” � Computational genomics in TAU ◦ Ron Shamir’s lab – focus on gene expression and regulatory networks ◦ Eithan Ruppin’s lab – focus on metabolism ◦ Tal Pupko’s and Benny Chor’s labs – focus on phylogeny ◦ Roded Sharan’s lab – focus on networks ◦ Noam Shomron’s lab – focus on mi. RNA ◦ Eran Halperin’s lab – focus on genetics

“Solved” problems � Alignment � Protein coding gene finding � Assembly of long reads � Basic microarray data analysis � Mapping of transcriptional regulation in simple organisms � Functional profiling in simple organisms

“Worked on” problems � Determining protein abundance � Assembly of short reads � Transcriptional regulation in higher eukaryotes � “Histone code”: Chromatin modifications, their function and regulation � Functional profiling of mammalian cells � Association studies for single-gene effects � Construction and modeling of synthetic circuits

“Future” problems � Digital gene expression from RNA-seq studies � Prediction of nc. RNAs and their function � Global mapping of alternative splicing regulation � Integration of multi-level signaling (TFs, mi. RNA, chromatin) � Association studies for combinations of alleles

Using sequencing to find new antibiotics � All microbial genomes are sequenced in E. coli � Each sequencing efforts basically introduces genes (3 -8 Kb fragments) into E. coli � Sometimes sequencing fails � Idea: sequencing fails barrier to horizontal gene transfer

Using sequencing to uncover structural variation � Even sequencing of reads with 100 s of bp will no identify many indels � Idea: sequence pairs of sequences at some distance apart from each other

Mutational landscape of human cancer � High-throughput sequencing can identify all the mutations in different cancers � 20, 857 transcripts from 18, 191 human genes sequenced in 11 breast and 11 colorectal cancers.

Mutational landscape of human cancer few mutations are drivers most are passangers � Most studies did not identify high frequent risk allels � But: members of some pathways are affected in almost any tumour � Network biology needed � Problems:

Predicting nc. RNAs � Using histone modifications and sequence conservation to uncover long noncoding RNAs (linc. RNA)

Using conservation to uncover regulatory elements � 12 fly species were sequenced to identify ◦ Evolution of genes and chromosome ◦ Evolutionary constrained sequence elements in promoters and 3’ UTRs � Starting point – genome-wide alignment of the genomes