Choice of probes and restriction enzymes to generate RFLP data for the calculation of inter-line genetic distances and for the presentation of associations among inbred lines

--O.S. Smith, J.S.C. Smith, S.L. Bowen, R.A. Tenborg

Thirty-seven inbred lines that represent a broad array of diversity for that utilized in the U.S. Corn Belt and which included pairs of lines related by pedigree from 0% to 95% (Malecot's Coefficient of Parentage) were profiled by RFLPs for 157 probes and 257 combinations of probes by restriction enzyme digests of genomic DNA. Pedigree data and listings of probes have been given by Smith et al., (Theor. Appl. Genet. 80:in press, 1990). Distances were calculated between lines (Nei and Li, Proc. Natl. Acad. Sci. USA 76:5256-5273, 1979) from RFLP profiles for all probe/restriction enzyme combinations that were run and also for the following subsets of probes or restriction enzyme digests: 1) probes versus BamHI digests; 2) probes against EcoRI digests; 3) probes against HindIII digests; 4) only probes that showed >10 variants across the 37 lines; 5) Pioneer Hi-Bred International, Inc., probes only; 6) University of Missouri (UMC) probes only; and 7) Brookhaven National Laboratory (BNL) probes only. Correlation coefficients of pairwise distances between lines for different sets or subsets of probes and restriction enzymes ranged from a low of r = 0.85 (UMC versus BNL probes) to a high of r = 0.994 (complete set of probe enzymes versus probe/enzymes showing >10 variants). Correlation coefficients for different restriction enzyme combinations versus the complete dataset were r = 0.96 (BamHI), r = 0.96 (EcoRI), and r = 0.97 (HindIII). Correlation coefficients for different sources of probes versus the complete dataset were r = 0.94 (BNL), r = 0.95 (UMC), and r = 0.98 (Pioneer). If a subset of probes was to be chosen in order to limit the number of gels and hybridizations that would need to be made for a survey of diversity among lines then use of probes and restriction enzymes that reveal most polymorphism would be recommended on the basis of the results obtained from these data. It would be important that any set or subset of probes used to generate profiles for the showing of associations between lines or hybrids be selected so that probed sites are well dispersed throughout the genome and thus can provide a thorough sampling of the genome.

Return to the MNL 65 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage