BSA and Variant Analysis

Bulk segregant analysis (BSA) of a Plasmodium falciparum genetic cross (MAL31 × KH004). The R workflow computes reference-allele frequencies for each bulk sample from read depths, drops low-depth sites, removes local outliers with a sliding-window MAD filter, and smooths the frequencies along each chromosome.

Contact: Nirjhar Bhattacharyya (nbhattac@nd.edu)
Code: https://github.com/NirjharBhattacharyya/Bulk_Segregation_Analysis_Ferdig_Lab. The repository is private, ask Nirjhar for access. The analysis is in BSA_OPTIMISATION_MAL_KH.
Data: data/BSA6.1.multi.MAL31xKH004.SNP.filter.table (26.7 MB), a tab-separated SNP table with 12,803 variants on 14 chromosomes and allele and read depths for 114 bulk samples.

Minimal run

Clone the code into code/upstream (see code/README.md). Then, once, from this directory on a CRC front end:

module load R/4.4.0/gcc/11.5.0
mkdir -p .R-library results
export R_LIBS_USER=$PWD/.R-library
Rscript -e 'install.packages(c("dplyr","magrittr","locfit","reshape2","ggplot2"), lib=".R-library", repos="https://cloud.r-project.org")'
# The current Deriv (needed by doBy) requires R 4.5, so install an older one.
Rscript -e 'install.packages("https://cloud.r-project.org/src/contrib/Archive/Deriv/Deriv_4.1.6.tar.gz", lib=".R-library", repos=NULL, type="source")'
Rscript -e 'install.packages("doBy", lib=".R-library", repos="https://cloud.r-project.org")'

Submit from crcfe01 or crcfe02:

qsub scripts/run-bsa.sge

BSA_OPTIMISATION_MAL_KH is a saved interactive R session and can’t be run as is. scripts/run-bsa.R runs its functions with the original parameters as a batch job, writing to results/baseline/. It stops before the QTLseqr G′ section, which needs a package and inputs that aren’t included. For HTCondor, use scripts/run-bsa-condor.submit instead.