wilcas / sex_specific_mqtl

Sex-dependent placental mQTL provide insight into the prenatal origins of childhood-onset traits and conditions

Code and notebooks corresponding to analysis done in the above titled publication ¹. mQTL summary statistics and related annotations can be found on this Open Science Framework site.

Directory structure

• NICHD quality control for DNA methylation and genotyping
• RICHS quality control for DNA methylation and genotyping
• Analysis code for publication
  • main results figures and data processing
  • Supporting R scripts
  • Supporting TORQUE/MOAB scripts, written in bash

Environment file

All packages used in this project are included in the environment.yml file. They can be installed using anaconda

Footnotes

1. William Casazza, Amy M. Inkster, Giulia F. Del Gobbo, Victor Yuan, Fabien Delahaye, Carmen Marsit, Yongjin P. Park, Wendy P. Robinson, Sara Mostafavi, Jessica K. Dennis. Sex-dependent placental mQTL provide insight into the prenatal origins of childhood onset traits and conditions. iScience, 2024. https://doi.org/10.1016/j.isci.2024.109047 ↩

sex_specific_mqtl's Issues

create annotation files for each mQTL annotation and integrate

can use baseline model plus my annotation if it matches and there's a comma separating this file and the baseline in the `--ref-ld-chr` command

---

sex_specific_mQTL/delahaye_mQTL_analyses/pbs_scripts/run_heritability.pbs

Lines 12 to 17 in 4323c07

	# @todo create annotation files for each mQTL annotation and integrate
	# @body can use baseline model plus my annotation if it matches and there's a comma separating this file and the baseline in the `--ref-ld-chr` command
	ldsc() {
	/arc/project/st-dennisjk-1/software/ldsc/ldsc.py --h2 /scratch/st-dennisjk-1/wcasazza/tmp_GWAS/formatted_sumstats/${1}.sumstats.gz \
	--ref-ld-chr /scratch/st-dennisjk-1/wcasazza/1000G_phase3_ldsc/partitioned_1kg_common/merged_sex_interaction_marginal.\
	--w-ld-chr /scratch/st-dennisjk-1/wcasazza/1000G_phase3_ldsc/weights_1kg_common/1kg_cosmopolitan.common.\

---

This issue was generated by todo based on a todo comment in 4323c07. It's been assigned to @wilcas because they committed the code.

implement QTL Quality filters -->

delahaye et al 2018 use FDR based on 1000 permutations to narrow down their initial number, they then restrict 5% of each allelic dosage. I should report something at least including the dosage (especially in the sex specific hits) and perhaps a more stringent cutoff like bonferroni or a lower FDR. Another thing I should do is report those mQTLs with non-zero CPP in CAVIAR -->

---

sex_specific_mQTL/delahaye_mQTL_analyses/post_mqtl_analyses.Rmd

Lines 109 to 113 in 96c2a4b

	<!-- @todo implement QTL Quality filters -->
	<!-- @body delahaye et al 2018 use FDR based on 1000 permutations to narrow down their initial number, they then restrict 5% of each allelic dosage. I should report something at least including the dosage (especially in the sex specific hits) and perhaps a more stringent cutoff like bonferroni or a lower FDR. Another thing I should do is report those mQTLs with non-zero CPP in CAVIAR -->
	## TESTING
	Checking if these mQTLs include too many low-frequency snps:
	```{r}

---

This issue was generated by todo based on a todo comment in 96c2a4b. It's been assigned to @wilcas because they committed the code.

generate annotation for aries cordblood mQTL

The database provides all mQTLs with FDR > 0.05 as well as a map of SNPs and probes. The format is MatrixEQTL, so my code should more or less apply, I will just need to change and specify paths etc. What I need to do is:

• finemap with CAVIAR
• Generate annot files
• re-run ldsc with this annot appended
• CHECK if these have sex chromosomes

Edit this for final threshold table\n",

sex_specific_mQTL/delahaye_mQTL_analyses/heritability_check.ipynb

Lines 196 to 201 in bb1b773

	" }else if (annot == \"all_cis\"){ # @todo Edit this for final threshold table\n",
	" merged <- merge(\n",
	" baseline_annot,\n",
	" tmp[,.(all_cis=ifelse(any(FDR < 0.05),1,0)),by=\"SNP\"],by=\"SNP\",all.x=TRUE,sort = FALSE)\n",
	" merged[, `:=`(all_cis = ifelse(is.na(all_cis),0,all_cis))]\n",
	" return(merged[,.(CHR,BP,SNP,CM,all_cis)]) \n",

---

This issue was generated by todo based on a todo comment in bb1b773. It's been assigned to @wilcas because they committed the code.