We design chips that enable the detection of alternative splice variants of genes. The design optimally chooses segments representing the alternative splice variants of each gene. Probes are selected from each segment according to a list of criteria including specificity, distance from the 3' end, sequence quality, GC content, and so on. The designs are based on predictions made by the LEADS software platform that clusters and assembles ESTs, known mRNAs and genomic data. For each gene, the LEADS software produces a comprehensive list of mRNA transcripts, each of which is a different predicted splice variant. Sequence areas that are multiply covered are used to detect and eliminate sequencing errors. These areas are also used for the detection of polymorphisms, which can be used in genotyping chips. Having good designs is crucial to extract meaningful information from chip experiments. Designs that do not use all available data, splice variants and sequencing errors might lead to useless probes and misleading results hiding important biological phenomena. It is believed that at least 35% of human genes have alternative splice variants. Different splice variants may have different functions, be expressed in different tissues or indicate a disease. Therefore, it is important to distinguish between the expression patterns of different splice variants. This is achieved, using LEADS-derived data, by choosing probes that are unique to some of the variants. If one just wishes to measure the overall expression level of the gene, probes that are common to all the variants can be chosen.
Copyright. All Rights Reserved.