This entry contains the scripts for genetic analyses of our article "Genetic mechanisms for impaired synaptic plasticity in schizophrenia revealed by computational modelling"
Gene-expression analysis
The population-averaged SCZ-versus-HC protein concentration changes and ion-channel conductance changes are all included in the simulation scripts of folders syn and l23pc.
The subject-wise effects on model parameters cannot be included in this repository due to data distribution restrictions.
However, the raw data are freely available to anyone who registers, signs the agreement of distribution and is approved by NIMH repository & genomics resource.
Here, we include all the details on how to process the raw data into the factors of model parameters used in the paper.
First, download the data and run the gene-expression analysis using the R script
run_commonmind_normalization.R
for both ACC and PFC data - follow also the instructions to get the neuron-imputed data mentioned in the comments of the script.
Then, run the scripts extract_commonmind_data_ACC.py
and extract_commonmind_data_PFC.py
that analyze the data and save the subject-wise coefficients for protein concentration and ion-channel conductance parameters.
Genotype-phenotype analysis
To make the lists of SNPs in genes of interest, run the following commands:
sh dofilterbegends.sh #Looks for the chromosome numbers and start and end base pairs of each gene of interest. The file ref_GRCh37.p13_top_level.gff3, which is not included in this entry due to large size but is freely available, is needed for this.
for q in genes_new genes_new_synaptic genes_all_synaptic_PKA genes_all_synaptic_PKC genes_new_ionchannels genes_lips_synaptic_ABFGJLOQ
do #This for loop runs the SNP finder for many gene sets. It needs the file PGC3_SCZ_wave3.primary.autosome.public.v3.vcf.tsv
python3 filter_ripke2020_v3_givenlist.py $q #which is publicly available at the PGC website (currently, it can be downloaded at https://figshare.com/articles/dataset/scz2022/19426775).
done #This takes some 5-20 min per gene group to run.
sh rungenes.sh #This is the main script for gene association analysis. However, sensitive data were used for this that are not included here. The script outputs .best files
#that contain the PRS scores for each subject - these values were correlated to the EEG indices (also sensitive data and excluded from here) and used for drawing Fig. 5.
python3 drawcorrelations_PRS_EEGphenotype.py
python3 drawcorrelations_PRS_EEGphenotype_suppl.py