1a.Objectives (from AD-416)
Develop statistical, quantitative genetic, genomic, biometric, and computing technologies for use in combining genotypic information from DNA from bulls provided by artificial insemination organizations with phenotypic data on yield and fitness traits.

1b.Approach (from AD-416)
Research is directed at developing genomic tools to estimate effects of chromosome segments or individual genes. Previous joint research have used widely spaced markers and found loci with large effects, but confirmed that genetic variation for most traits is quantitative, with many loci distributed across the genome individually exerting small effects. Planned research will estimate the trace inheritance of the many smaller genetic effects by using available single nucleotide polymorphisms (SNPs).

Research is directed toward predicting transmitting abilities (PTA) more accurately by simultaneously predicting effects of many genetic loci using dense marker genotypes. Phenotypic data used for this will be from the national genetic evaluations for traits currently evaluated. Data will be expressed as daughter yield deviations or as de-regressed PTAs to avoid reanalysis of all raw records. Use of dense markers to estimate chromosome segment effects avoids most of the recombination between markers and causative genes that reduced the accuracy of previous within-family analyses. Predictions for young bulls will be more reliable when DNA samples for these are obtained in a later phase of the research.

3.Progress Report

The project is related to in-house subobjective 3a (develop methodology for calculation of genome-enhanced breeding values using SNP genotypes). Software tools were designed to analyze large numbers of single-nucleotide polymorphisms (SNPs) efficiently for association with quantitative traits and to calculate genomic breeding values. The MMAP (mixed model analysis in pedigrees) program integrates established and novel algorithms to analyze SNP data in pedigrees and allows genomewide association analysis using single or multi-SNP and/or haplotype fixed effect models to account for residual variation through a relationship matrix derived from pedigree or genomic data. The MMAP program will be distributed as free software. A software program to split genotypes into haplotypes is being developed as a step toward imputation of missing SNP genotypes. Visualization methods for results from genomic predictions were examined for their ability to present data with higher density than possible with text or tables, provide additional insight into data, and allow data exchange without disclosure of sensitive information. Genetic variance ratios (actual:expected) can be plotted as bar graphs to allow easy identification of chromosomes that deviate from expectation; stacked bar graphs allow for simultaneous comparison of estimation methods for variance ratios. All markers affecting a trait can be plotted on the same ordinate to visualize distribution of marker effects across the genome; colors or textures can be used to differentiate between chromosomes; and stacked graphs can be constructed to compare trait groups. Chromosomal breeding values can be presented as sparklines (high-resolution graphics embedded in text) to provide an overview of individual animals for comparison to potential mates. Small multiples of chromosomal genetic correlation matrices can be used with edge exclusion graphs to identify association patterns among traits. Line plots of marker effects for autosomal recessives can be used to locate chromosomal regions with probable causative mutations quickly. Such graphics are easily produced automatically and can be added to online query systems to provide users with novel information at little cost. Three presentations (2 oral, 1 poster) were made at the 2009 meeting of the Federation of Animal Science Societies; 3 abstracts were published in the meeting proceedings. Two proceedings papers were published by the International Committee on Animal Recording; 2 scientific papers were published by the Journal of Dairy Science. The project was extended to May 31, 2010, and funding was increased to allow development of haplotyping algorithms as well as efficient methods and software for genetic prediction and parentage verification using low-density SNP chips and estimation of multibreed SNP effects. Monitoring activities for the project included discussion in Beltsville, MD, of project plans, goals, and accomplishments with the Cooperator’s principal investigator, who has an office at the Animal Improvement Programs Laboratory to allow frequent, direct interaction with Laboratory researchers.