Diversity of microsatellites derived from genomic libraries and GenBank sequences in rice (Oryza sativa L.)

The growing number of rice microsatellite markers warrants a comprehensive comparison of allelic variability between the markers developed using diffe rent methods, with various sequence repeat motifs, and from coding and non- coding portions of the genome. We have performed such a comparison over a s et of 323 microsatellite markers; 194 were derived from genomic library scr eening and 129 were derived from the analysis of rice-expressed sequence ta gs (ESTs) available in public DNA databases. We have evaluated the frequenc y of polymorphism between parental pairs of six inter-subspecific crosses a nd one inter-specific cross widely used for mapping in rice. Microsatellite s derived from genomic libraries detected a higher level of polymorphism th an those derived from ESTs contained in the GenBank database (83.8% versus 54.0%). Similarly, the other measures of genetic variability [the number of alleles per locus, polymorphism information content (PIC), and allele size ranges] were all higher in genomic library-derived microsatellites than in their EST-database counterparts. The highest overall degree of genetic div ersity was seen in GA-containing microsatellites of genomic library origin, while the most conserved markers contained CCG- or CAG-trinucleotide motif s and were developed from GenBank sequences. Preferential location of speci fic motifs in coding versus non-coding regions of known genes was related t o observed levels of microsatellite diversity. A strong positive correlatio n was observed between the maximum length of a microsatellite motif and the standard deviation of the molecular-weight of amplified fragments. The rel iability of molecular weight standard deviation (SDmw) as an indicator of g enetic variability of microsatellite loci is discussed.