Skip to content

MAGMA HBA Analysis

I applied MAGMA to the PGC 2022 GWAS of schizophrenia using single-cell RNAseq read-counts from the Human Brain Atlas1 as reference data. This was an attempt to verify that I could reproduce the results of Duncan et al.2.

Results

The results are plotted below:

sch-human-brain-magma

These results closely match those of Duncan et al. Like Duncan et al., we find that the most significant cell type for Schizophrenia is the Somatostatin MGE Interneuron.

When we apply a technique based on Watanabe et al.'s to identify independently significant clusters, we get the following results:

Retained_clusters P Supercluster Class auto-annotation Neurotransmitter auto-annotation Neuropeptide auto-annotation Subtype auto-annotation Transferred MTG Label Top three regions Top Enriched Genes
Cluster239 5.4105e-16 MGE interneuron NEUR GABA CCK CHGA CHGB CORT IGF NAMPT NMU NPPC NUCB NXPH PNOC SCG SST TAC UBL VGF proSAAS INT-SST Cerebral cortex: 95.4%, Hippocampus: 1.5%, Basal forebrain: 1.2% AC022905.1, NMU, MAFB, SCARA5, PRDM1, TRBC2, MYO5B, PDCL2, BLM, PNOC
Cluster280 1.8714e-15 CGE interneuron NEUR GABA CCK CHGA CHGB CRH IGF NAMPT NUCB PNOC SCG UBL VIP proSAAS INT-VIP Cerebral cortex: 73.7%, Amygdala: 13.3%, Hippocampus: 7.8% AC006305.1, LINC01416, AL391832.4, LINC01905, SP8, DCN, CHRNB3, CXCL14, CHRNA7, CRH
Cluster132 2.4393e-13 Miscellaneous NEUR VGLUT1 VGLUT2 CBLN CCK CHGA CHGB NAMPT NUCB SCG UBL VGF proSAAS 0 Cerebral cortex: 98.4%, Hippocampus: 0.8%, Midbrain: 0.8% CD53, DPP4, AC060765.1, LINC01721, AC112770.1, AL157769.1, VIPR2, PCSK1, AC016687.2, QRFPR
Cluster231 1.9584e-12 Eccentric medium spiny neuron NEUR GABA CHGB NAMPT NUCB PDYN PENK SCG VGF proSAAS MSN-D1 Amygdala: 60.2%, Basal forebrain: 16.4%, Cerebral cortex: 13.7% LMNTD1, NPFFR2, AC012078.2, ZNF736P9Y, GABRQ, EYA2, PCDH11Y, CAPSL, DRAXIN, ITGA6
Cluster404 6.6585e-08 Miscellaneous NEUR VGLUT1 CBLN CCK CHGA CHGB IGF NAMPT NUCB NXPH SCG SST TAC VGF proSAAS 0 Cerebral cortex: 94.4%, Amygdala: 2.5%, Basal forebrain: 1.2% AC016687.2, AC021134.1, LINC02378, LINC02263, LINC01915, CXCL14, AL033539.2, AC004862.1, AC073578.2, LINC02306
Cluster407 7.0269e-06 Amygdala excitatory NEUR VGLUT2 VGLUT3 CART CCK CHGA CHGB NAMPT NUCB NXPH SCG TAC UBL VGF proSAAS 0 Amygdala: 72.6%, Cerebral cortex: 23.5%, Hippocampus: 2.7% TFAP2C, PAPPA2, CARTPT, BARHL2, LINC01798, MEIS1-AS3, AC006065.4, LHX9, GALR1, POU2F3
Cluster456 1.6412e-05 Thalamic excitatory NEUR VGLUT2 ADCYAP CART CBLN CCK CHGA CHGB CRH NAMPT NUCB NXPH SCG UBL VGF proSAAS 0 Thalamus: 99.8%, Hippocampus: 0.2%, Cerebral cortex: 0.0% ST8SIA6, AC012535.1, AC093334.1, ST8SIA6-AS1, AP000331.1, AC006487.1, SHOX2, ANXA3, AC087482.1, VANGL1

Here is a the plot of significant clusters again, this time with independently significant clusters labeled:

sch-hba-cluster-plot

It is interesting to observe the fairly significant overlap with the significant clusters found in the analysis of the educational attainment GWAS.

  1. Kimberly Siletti, Rebecca Hodge, Alejandro Mossi Albiach, Ka Wai Lee, Song-Lin Ding, Lijuan Hu, Peter Lönnerberg, Trygve Bakken, Tamara Casper, Michael Clark, and others. Transcriptomic diversity of cell types across the adult human brain. Science, 382(6667):eadd7046, 2023. URL: https://www.science.org/doi/abs/10.1126/science.add7046. ↩

  2. Laramie E Duncan, Tayden Li, Madeleine Salem, Will Li, Leili Mortazavi, Hazal Senturk, Naghmeh Shahverdizadeh, Sam Vesuna, Hanyang Shen, Jong Yoon, and others. Mapping the cellular etiology of schizophrenia and complex brain phenotypes. Nature Neuroscience, 28(2):248–258, 2025. URL: https://www.nature.com/articles/s41593-024-01834-w. ↩