2014.02.12

• イベント

第46回インシリコ・メガバンク研究会開催のお知らせ（2月21日）

第46回インシリコ・メガバンク研究会を下記のとおり行いますのでご案内いたします。

今回は産業技術総合研究所 生命情報工学研究センター・Martin Frith先生を講師としてお迎えし、「元来の統計的アライメントにおける最新高スループットDNAの適応」について講演していただきます。

日時：平成26年2月21日(金)　17：00‐18：30
場所：東北メディカル・メガバンク機構2階会議室1

演題：Adapting classic statistical alignment to modern high-throughput DNA
講師：Dr. Martin Frith（産業技術総合研究所 生命情報工学研究センター）

＊本講演は医学系研究科系統講義コース科目の授業として振替可能です。

・要旨： For many decades, the main way of analyzing biological sequences has been by comparing and aligning them.  This remains true today.  Modern tasks include: comparing whole genomes; aligning bisulfite-converted DNA reads to a genome; aligning long, high-error sequences from single molecule sequencers; aligning ancient or degraded DNA; comparing metagenomic DNA to a protein database. Over the decades, statistically powerful alignment techniques have been developed, including: log likelihood ratio scoring matrices, pair hidden Markov models, and probabilistic alignment.  Unfortunately, these methods are rarely used with modern deep sequencing data, perhaps because they are thought to be too slow.  This talk will demonstrate that they can be made fast enough, and offer great benefits.　This talk will also sketch how to incorporate some new features into classic statistical alignment: sequence quality data (phred scores); pairing relationships between DNA reads; and split alignment, where different parts of one query sequence may align to disjoint loci in a genome.  This is useful for genome rearrangements, spliced RNA (including trans-splicing), and even whole genome comparison.  The statistical approach can tell us the reliability (unambiguity) of any part of an alignment.

・世話人: 長﨑正朗