Epsilon Archive for Student Projects

Abstract

The domestication process and artificial selection can lead to an increased proportion and frequency
of deleterious genetic variants affecting health and productivity in cattle. The aim of this project was
to estimate the load and allele frequency spectrum of different kinds of functional and possibly
deleterious variants (missense, potentially loss-of-function, potentially gene-regulatory) in 3
different cattle breeds (Brahman, Hereford, and Holstein), predict with bioinformatic tools the effect
of these variants on proteins, and investigate the pairwise linkage disequilibrium between
deleterious variants, and between deleterious variants with their surrounding area.
For the analysis the 1000 Bull Genome dataset was used. Deleterious missense and loss-of-function
variants were identified using MutPred2 and MutPredLOF, respectively, and Transcription Start
Site and enhancer variants were identified using Cap Analysis Gene Expression sequence data.
PLINK was utilized for the allele frequency and linkage disequilibrium computations. The
performance of MutPred2 was evaluated by performing MutPred2 analysis on deleterious variants
present in the Online Mendelian Inheritance in Animals (OMIA) database.
The results showed low deleterious mutation load and an enrichment of deleterious variants at low
frequencies, indicating the effectiveness of purifying selection at purging them from the population.
Moreover, balancing selection may be a potential mechanism for the higher frequencies observed
for a small amount of deleterious variants. The observed breed-specific differences in load and allele
frequencies may be attributed to differences in effective population size, selection pressure, and
breeding strategies. Despite lacking proper annotation, the enrichment of regulatory variants at
lower frequencies suggests that they are under the influence of selection. The observed low linkage
disequilibrium between and around highly deleterious variants may be attributed to their low
frequencies and their presence on different haplotypes due to recombination. Evaluation of
MutPred2 revealed that the inclusion of Position-Specific Iterative Basic Local Alignment Search
Tool (PSI-BLAST) in the analysis is necessary for the reliable identification of deleterious variants.
This study, if implemented accurately on a large scale, has the potential to facilitate the development
of a genomic database that can contribute to reducing the frequencies of deleterious variant through
genomic selection and improving predictions on causative variants linked to genetic defects,
ultimately enhancing the effective management of genetic diseases.