Scripts for the manuscript:

Network-wide risk convergence in gene co-expression identifies reproducible genetic hubs of schizophrenia risk

Data repository

• Zenodo: ANCR scores, connectivity to PGC3 ranks, and other data
• Data pertaining to certain scripts may also be found in their respective sub-directories

Computing ANCR (Compute_ANCR, Figures 2,3,4):

• ANCR_step1_xGene_RemNetInMod.R : Compute network-wide risk convergence onto a module ("z_kme", gene risk vs gene kme), for all modules in a set of networks. (Fig 2B, 3B, 4).
• ANCR_step1_xGene_RemNetInNet.R : Compute network-wide risk convergence onto a module ("z_kme", gene risk vs gene kme), for all modules in a set of networks. (Fig 3D).
• ANCR_step2_xModule_RemNetInMod.R : Compute across module association between module median_MAGMA and z_kme, accounting for module size. This across module association is used as an index of ANCR.(Fig 2B, 3B, 4).
• ANCR_step2_xModule_RemNetInNet.R : Compute across module association between module median_MAGMA and z_kme, accounting for module size. This across module association is used as an index of ANCR. (Fig 3D).

Pseudobulk and pre-process published snRNA-seq data (snRNA_nets, Figure 4):

• Batiuk_pseudobulk.R] : Pseudobulk and preprocess snRNAseq data from Batiuk et al. (Fig 4).
• Batiuk_data : snRNA-seq_and_spatial_transcriptomics.zip. Paper : https://doi.org/10.1126/sciadv.abn8367
• Ruzicka_pseudobulk.R] : Preprocess pre-pseudobulked snRNAseq data from Ruzicka et al. (Fig 4).
• Ruzicka_data : pseudobulk_mean_logcounts_filtered.rds. Paper : https://doi.org/10.1101/2022.08.31.22279406
• snRNA_WGCNA.R : Compute networks per celltype in a snRNAseq dataset, using the pre-processed expression matrices constructed in the above scripts. (Fig 4).

Computing and evaluating gene connectivity to PGC3 risk loci genes (PGC3_connectivity, Figures 5,6):

• PGC3conn_Fig5.R : Compute ranked connectivity to PGC3 SCZ risk genes (and null sets), taking into account connectivity to background. (Fig 5).
• PGC3conn_CRISPRa_null_Fig5.R : Evaluate association between PGC3 connectivity (for real and null sets) and Zscore of response to CRISPRa activation of PGC3 eGenes (from Townsley et al). (Fig 5).
• Townsley.et.al_CRISPRaPGC3_data : Published data of Zscore of response to CRISPRa activation of PGC3 eGenes : see their Supplemental Data 1.
• acrosskme_genes_Fig6A.R : Compute per-gene connectivity to SCZ risk based on across-module association between kME and median magma. (Fig 6).

For any data or code inquiries please contact Giulio Pergola: [[email protected]] [https://www.libd.org/team/giulio-pergola-phd/]