Introduction to the Data Set

  • Leptinob/ob mice do now produce insulin and become obese due to overeating.
  • This study crossed mice carrying the Leptinob/ob mutation in C57BL/6J and BTBR T+ tf/J.
  • C57BL/6J mice are resistant to diabetes and BTBR mice are susceptible.
  • By crossing these two strains, the authors aimed to identify genes which influence susceptibility to T2D.

Input File Format

  • QTL mapping data consists of a set of tables of data: genotypes, phenotypes, marker maps, etc.
  • These different tables are in separate comma-delimited (CSV) files.
  • In each file, the first column is a set of IDs for the rows, and the first row is a set of IDs for the columns.
  • In addition to primary data, a separate file with control parameters (or metadata) in either YAML or JSON format is required.
  • Published and public data already formatted for QTL mapping are available on the web.
  • These data can be used as a model for formatting your own QTL data.

Calculating Genotype Probabilities

  • The first step in QTL analysis is to calculate genotype probabilities.
  • Calculate genotype probabilities between genotyped markers with calc_genoprob().

Performing a Genome Scan

  • A qtl2 genome scan requires genotype probabilities and a phenotype matrix.
  • The output from a genome scan contains a LOD score matrix, map positions, and phenotypes.
  • LOD curve plots for a genome scan can be viewed with plot_scan1().
  • A genome scan using sex as an additive covariate searches for QTL which affect both sexes.
  • A genome scan using sex as an interactive covariate searches for QTL which affect each sex differently.

Calculating A Kinship Matrix

  • Kinship matrices account for relationships among individuals.
  • Kinship is calculated as the proportion of shared alleles between individuals.
  • Kinship calculation is a precursor to a genome scan via a linear mixed model.

Performing a genome scan with a linear mixed model

  • To perform a genome scan with a linear mixed model, supply a kinship matrix.
  • Different mapping and kinship calculation methods give different results.
  • Using a set of Leave-One-Chromosome-Out kinship matrices generally produces higher LOD scores than other methods.

Performing a Genome Scan with Binary Traits

  • A genome scan for binary traits (0 and 1) requires special handling; scans for non-binary traits assume normal variation of the residuals.
  • A genome scan for binary traits is performed using logistic regression.

Finding Significant Peaks via Permutation

  • A permutation test establishes the statistical significance of a genome scan.
  • 1,000 permutations provides a good estimate of the significance threshold.

Finding QTL peaks

  • LOD peaks and support intervals can be identified with find_peaks().
  • The Bayesian Credible Interval estimates the width of the support interval around a QTL peak.
  • Using a higher prob value for the Bayesian Credible Interval results in a wider support interval.

Estimating QTL effects

  • Estimated founder allele effects can be plotted from the mapping model coefficients.
  • Additive and dominance effects can be plotted using contrasts.

Integrating Gene Expression Data

  • There will be many genes under a QTL peak.
  • You can search for genes with SNPs that produce coding changes by querying a VCF file.
  • You can search for genes with expression changes that may influence your phenotype by performing mediation analysis with expression data from the same mice.

QTL Mapping in Diversity Outbred Mice

  • There are generally five steps to QTL mapping in DO mice:
    • map the trait,
    • perform permutations,
    • find significant peaks,
    • calculate founder allele effects at the QTL peak,
    • perform association mapping to narrow the gene candidates.
  • You may need to bring in outside resources to help narrow your candidate gene list.
  • You will need the 10 GB SNP database to perform association mapping.