Bioinformatic Mining Of Microsatellite From Expressed Sequence Tags Of Cattle And Sheep

With the development of animal genomics, a great deal of valuable data is generated, of which EST is an important part. It has been an emphasis part of nowadays research to mine valuable information from these huge number EST data by bioinformation technique.In this study, SSRs markers were developed by bioinformatic methods. 41986 cattle and sheep UniGene sequences downloaded from the National Center for Biotechnology Information were mined for the identification of microsatellites. Cattle and sheep UniGene sequences were screened for the presence of perfect microsatellites by using SSRIT(online) and SSRFinder(unpublished) soft tools. All cattle sequences were 41986, while all sheep sequences were 4081. Cattle UniGene databases were used to identify and characterize SSRs using the Perl program SSRFinder developed for this study. This computer program was run under Linux. Using a FASTA-formatted sequence file containing multiple sequences, the SSRFinder was written to search each sequence for all possible combination of di-, tri-, tetra-, penta- and hexa-nucleotide repeats with the criteria of minimum number of repeats of 7 for dinucleotide, 6 for trinucleotide, 5 for tetranucleotide, 4 for pentanucleotide and 3 for hexanucleotide. Single nucleotide repeats were not selected because they are generally not considered as being useful as polymorphic markers. These UniGene sequences were screened for the presence of perfect microsatellites. In sheep sequences, a total of 136 SSRs were identified from 121 EST sequences.The frequency of EST containing SSR is 3.0%, which represent an average density of one microsatellites/22.47kb. In cattle sequences, a total of 1831 SSRs were identified from 1666 EST sequences.The frequency of EST containing SSR is 4.0%, which represent an average density of one microsatellites/19.89kb.In cattle, the dinucleotide repeat motif was the most abundant SSR, accounting for 54%, followed by 22%, 13%, 7% and 4%, respectively, for tri-, hexa-, penta- and tetra-nucleotide repeats.In sheep, trinucleotide repeat motif was the most abundant SSR, accounting for 40%, followed by 34%, 21%, 3% and 2%, respectively, for tri-, hexa-, penta- and tetra-nucleotide repeats. Among the dinucleotide repeats of cattle and sheep, AC/TG was the most abundant type and TA/AT was second,while GC/CG was the least type. Among the dinucleotide repeats.In cattle,AGC was the most abundant type, followed by CGG type, the third type was CTG.In sheep, CTG was the most abundant type, followed by CGG, the third type was AGG. In order to obtain an idea about putative functions of SSR-containing genes, similarity comparison using BLASTX revealed the identities of ESTs found here to contain microsatellites. These sequences were compared to the nonredundant (nr) protein database of the NCBI Database using 1e-05 as a cut-off expected value. This results indicated a total of 698 cattle sequences are for known genes, while 968 are still unknown genes. In sheep, a total of 45 sequences are for known genes, while 76 are still for unknown genes. 54 primer pairs were designed from sheep EST-SSRs using computer program Primer Premier (Version 5.00)and 31 primer pairs can show clear PCR products by electrophoresis for further polymorphism analysis. By AgNo3 was dyed, three primer pairs were presented polymorphism in in about 7 sheep species.In these three microsatellite loci, the alleles numbers distributed from 5,11and 17, and the range of these alleles were 170 to 301bp and similar with their predicted size. The expected heterozygosity in three microsatellite loci was from 0.771 to 0.841 and the PIC values distributed from 0.731 to 0.821. These values showed that these 3 microsatellite loci were suitable for genetic analysis.