比較陣列語言 Matvec與Octave於遺傳育種之應用

黃鈺嘉 林德育 廖仁寶 顏念慈 張秀鑾

行政院農業委員會畜產試驗所

應用陣列語言 Matvec與Octave 分別開發從選拔指數、混合模式、核酸序列分析DNA共有環帶分析的程式例集。 Matvec 為 Wang and Fernando (1995)於依利偌大學(UIUC)開發的遺傳育種用高階矩陣語言，Octave 高階矩陣語言則為GNU 共享軟體，目前二者均提供Window與 Linux 可執行版供教學與研究使用下載。由於Matvec 從數量遺傳學觀點出發，故提供如近親係數計算、一般線性模式計算及變方成份估算等多種傳統數量遺傳專用的功能，但Octave 較多的字元處理函數功能及 Matlab m-file 共容的特色，有利於核酸序列資料處理，且可共享已開發的生物計算程式。遺傳育種研究者可結合此兩種類似語言的優點以提昇工作效率。

關鍵語：矩陣、育種、遺傳。

COMPARSION OF MATVEC AND OCTAVE LANGUAGES FOR
GENETIC AND BREEDING PRACTICES

Y. C. Huang, D. Y. Lin, R. B. Liaw, N. T. Yen and H. L. Chang

Taiwan Livestock Research Institute, Council of Agriculture

For genetic and breeding practices, Matvec and Octave were used to develop program sets, which included programs of selection indices, generalized linear mixed model, genome sequence analysis and DNA bands summaries. Matvec was developed by Wang and Fernando (1995) at University of Illinois(UIUC). Octave was a GNU software. Matvec and Octave have both Window and Linux versions, which are free for research purpose. Because Matvec was developed from a quantitative geneticist point of view, therefore inbreeding coefficients, mixed models and variance components estimation were built-in elements. Octave with more functions for character and string operation is very handy for genome sequence analysis, and Octave is highly compatibility with m-files as of Matlab. Researchers can fasten together benefits of Matvec and Octave to achieve genetic and breeding goals.

Key Words: Matrix, Genetics, Breeding.