一种具有高度可验证性的微阵列片间校准算法

由于系统偏差的存在,多微阵列数据之间在进行差异比较分析前,首先需要进行片间校准,使不同微阵列的探针数据处于同一可比较的水平。目前片间校准算法的准确性依赖于看家基因表达水平的不变性,或者探针数据统计分布的稳定性。这些假设条件并非在所有条件下成立,如需要进一步实验验证,其算法的正确性又难以进行实验验证。本文提出一种具有高度实验可验证性的多微阵列数据校准的遗传选择优化方法。基于该方法所得到的优化校准基准子集,可以在仅假设少数看家基因表达水平不变的情况下,在不牺牲校准精度的前提下,完成多微阵列数据的片间校准,从而使得通过实验验证片间校准算法的正确性成为可能。

A Highly Verifiable Inter-Microarray Normalization Method

Due to the existence of systematic bias in the measurements, inter microarray normalization is required before the integration data analysis of multiple microarray datasets. The current inter microarray normalization methods rely on some assumptions, which rarely can be verified by experiments, such as the invariant expression level of housekeeping genes or the stability of statistical data distribution. In this paper, we present a highly verifiable inter microarray normalization method using a genetic optimization algorithm, through which only a small number of housekeeping genes are selected as the normalization markers to facilitate experiment validation without any precision lost.