By Invitation Only
Date: 29 July 2019, Monday
Time: 11am
Venue: MD6, Level 12 Conference Room (#12-02N)

Scientific Presentation:
Speaker: Touati Benoukraf
Title:   Building an Asian AML genome reference to improve RNA analysis 

Abstract:

A key step in RNAseq analysis is mapping sequenced reads to a genome reference. However, for certain diseases such as Acute Myeloid Leukemia, the human genome reference can be far from the actual patients’ genome of affected cells, due to the numerous structural variants (translocations, insertions, duplications, and deletions). More distant is the actual genome from the reference genome, less specific will be the mapping, and consequently, less accurate will be the final analyses and interpretations. The recent advent of third-generation sequencing (3GS) technologies brings promise for better characterization of genomic aberrations by virtue of having longer reads. Although current applications of 3GS are restricted by their high sequencing error rates and low sequencing throughput, we were able to overcome these limitations, by developing an algorithm that employs split-reads and hard-clipped reads for SV detection, combined to a neural network classifier for true SV enrichment on low-depth 3GS datasets generated by the Oxford Nanopore technology. In simulated data, we demonstrated that our method provides the highest SV detection accuracy (F1 score = 0.91) amongst other long-read SV callers using 12 gigabases (4X) of sequencing data. In patient samples, besides the detection of genomic aberrations, our method also uncovered many normal alternative sequences or alleles which were present in healthy individuals. The low sequencing depth requirements for our method enable the use of Nanopore sequencing for accurate SV characterization and understanding their effects on transcription regulation at a lower sequencing cost, an approach compatible with clinical studies.