Software for Statistical Analysis of Experimental Data
(Mixed Linear Model Analysis, Microarray Data Analysis and Diallele Cross Analysis)
Copyright 2007-2008 ©
Jian Yang, Junqi Cheng, Yangyun Zou, Zhen Xia and Jun Zhu
Institute of Bioinformatics, Zhejiang University, Hangzhou, China
This software can be freely redistributed with non-commercial purpose and under the condition of no changes to the software and its related documents. No warranty of any sort is provided for use of the software.
QTModel is user-friendly computer software which packaged with modules for microarray data analysis, diallele design analysis and mixed model analysis.The mixed model module is developed for analyzing data from experimental designs with random factors. It is now available for commonly used randomized block design, randomized complete block design, latin square design, factorial design, multi-factor factorial design, nested design, and cross nested design etc. For fixed factors, pair-wised comparisons are done for all possible pairs of fixed effects of one factor. For random factors, some mixed linear model approaches, such as MINQUE, MIVQUE, REML and EM, will be used to estimate the variances of these random factors, and also unbiased prediction methods, such as BLUP, LUP and AUP, are used to predict the random effects of the random factors.The module of microarray data analysis is based on the HAB-Array method (Henderson III and Bayesian method for microarray analysis), which is proposed for detecting DEGs (Differentially Expressed Genes) under specific treatments. Currently, it has the capacity of analyzing data with one or two treatment factors. If the treatment is treated as random factor, the effect of each treatment level can be predicted by MCMC (Markov Chain Monte Carlo) algorithm or conventional mixed linear model approaches such as BLUP (Best Linear Unbiased Prediction) and LUP (Linear Unbiased Prediction). While if the treatment is treated as fixed factor, pair-wise comparison will be done between all possible pair of treatment levels. The selected DEGs will be exported for cluster analysis by the software ClusterProject. The software is programmed by C/C++ programming language. The GUI (Graphic User Interface) of QTModel is developed by QT. Since QT is a cross-platform C++ class library, any program developed based on QT can be complied and run any operating system with QT installed. Now QTModel have binary executive programs available for Windows XP, RedHat Linux, and IBM AIX. We will describe the application of QTModel in the following section.
QTLModel-0.70Beta-Setup.exe (Last update 03/21/2008)
QTModel-UserManual.pdf (Last update 03/21/2008)
Error Messages and Troubleshooting
If you have any question or comment with our software, please contact to Prof. Jun Zhu, Institute of Bioinformatics, Zhejiang University, Hangzhou, Zhejiang 310029, China. Email: firstname.lastname@example.org.