Open NagiYoshitsugu opened 11 months ago
Fig. 1 修正可能な異常の例 (A-C)Human Body Mapデータセットの腎臓サンプル. (A) 赤と青の点はSAD修正前の観測分布と期待値分布であり,青い影は期待値分布推定との標準偏差である.期待値分布とは,定量器によって推定された期待値分布からガウス誤差の平均を差し引いたものである.各ポイントは転写産物に沿った50-bpのビンである.異常転写産物ENST00000545682.5はその最初のスプライシングジャンクション(上)の後に発現不足があり赤枠で示す.別の転写産物が修正に関与する(下). (B)SAD修正後の同じ転写産物ペアの分布. (C)2つの転写産物のタンパク質ドメインアノテーション.発現低下異常領域を赤枠で示す. (D-F) GEUVADISデータセットからのサンプル(アクセッションERR188088) (D)上段の転写産物ENST00000532808.5は修正可能な異常であることが同定され,その発現不足異常領域が赤枠で示されている.下段の転写産物はリードの再配置が関連する. (E) SAD修正後の観察分布と期待値分布 (F) 以前の転写産物のタンパク質ドメインアノテーション.赤枠で示したのは発現不足の異常領域. Fig. 2 修正不可能な異常の例と特徴 (A, B)遺伝子UBE2Q1の修正不能異常の例(A) と遺伝子LIMD1の修正不能な異常の例(B).どちらの例も Human Body Map データセットの心臓サンプルに見られる.赤と青の点は,異常転写産物の観察されたカバレッジ分布と期待されるカバレッジ分布.青の影は期待される分布推定の標準偏差,赤枠は発現不足の異常領域を示す.両遺伝子とも5′末端付近の転写産物領域は過剰発現し,3′末端付近の領域は過小発現している. (C)共通の異常転写産物の過剰発現領域と過小発現領域の位置.赤い対角線は,過剰(過小)発現領域が主に転写産物の前半(5′半)と重なる異常と,主に転写産物の後半(3′半)と重なる異常を区別している.ほとんどの異常について,過剰発現領域は主に異常転写産物の前半部分と重なり,過小発現領域は主に異常転写産物の後半部分と重なっている. (D) すべてのサンプルに共通する過小発現領域のエクソンの個数.y軸は全46サンプルで合計した数.過小発現領域は通常1つまたは部分的なエクソンしか含まない.
一言でいうと
リードカバレッジが期待値から大きく外れている領域(=発現異常)を検知する方法,および異常の原因を推定する方法を開発した.
論文リンク
https://doi.org/10.1016/j.cels.2019.10.005
著者
Cong Ma, Carl Kingsford
ジャーナル/会議名
Cell Systems
出版日(yyyy/MM/dd)
2019/11/27