Table 1 Top 40 risk genes for SLE identified by random forest prediction using Immunochip genotype data from SLE patients and controls.

From: Novel risk genes for systemic lupus erythematosus predicted by random forest classification

Predicted genes¹	Gene importance score⁴	Association with autoimmune diseases in the GWAS catalog⁵	Differential expression in B and T cells^8,9
BLK	51.8	SLE, RA, KD, pSS	B > T***
CLEC16A	49.2	SLE, IBD, UC, T1D, MS, CD, Psoriasis	B > T
STAT4	39.8	SLE, UC, IBD, RA, CD, pSS, Celiac, PBC	T > B***
ETS1	33.5	SLE, RA, Celiac, Psoriasis	T > B
ZNF804A	33.4	New	B > T***
ANK3 ³ CDK1 ²	33.0	New	T > B
BANK1	30.5	SLE, IBD, CD	B > T***
PSMG1	27.5	IBD, UC, CD, AS	T > B
TNIP1 ³	27.2	SLE, IBD, Psoriasis	B > T
PLEKHH2 THADA ²	26.7	SLE⁶, CD, IBD, MS	Low
TPI1P2 TNPO3 ² IRF5 ²	26.6	SLE, PBC, pSS	Low
IKZF1 ³	25.9	SLE, IBD, CD, UC	T > B
PTGER4	24.4	IBD, CD, UC, AS, MS	T > B
CD44	23.9	SLE, Vitiligo	T > B
IRF5	23.1	SLE, UC, IBD, RA	B > T***
IL2RA	22.8	SLE^6,7, IBD, CD, T1D, RA, MS, Vitiligo	T > B
TNFSF4	21.6	SLE, CD, RA, Celiac, MS	Low
SLC15A4	20.4	SLE	B > T
IL12A-AS1 IL12A ^2,3	20.1	SLE⁶, Celiac, PBD, MS, pSS	Low
HIP1	19.7	SLE	B > T
XKR6	18.2	SLE	Low
CPEB4	17.8	IBD, CD	B > T*
ZNF365 EGR2 ²	17.6	IBD, CD, UC, RA	T > B*
THADA	17.2	SLE⁶, IBD, CD, MS	B > T
GLIS3 RFX3 ²	16.7	T1D	Low
NCF2 ³	16.7	SLE	B > T
PHRF1 ³	16.6	SLE	T > B
PAPOLG	16.4	SLE⁶, CD, RA, Psoriasis	T > B
IL1R1	16.4	IBD, UC, CD, AS	Low
LRRK2	16.1	IBD, CD, UC	B > T***
UBAC2 GPR183 ²	15.8	IBD, CD	T > B
ZFP36L2 THADA ²	15.4	SLE⁶, IBD, CD, MS	T > B
PVT1	15.4	SLE⁶, RA, MS	T > B
ZMIZ1	15.3	IBD, CD, MS, Vitiligo, Psoriasis	Low
ELMO1	14.9	CD, RA, PBC, Psoriasis	B > T
WDFY4	14.9	SLE, RA	B > T***
AKAP11 TNFSF11 ²	14.8	IBD, CD	B > T
DOCK3 MANF ²	14.7	New	Low
SATB2	14.7	UC, IBD	Low
IRF8	14.6	SLE, IBD, UC, RA, PBC, CD	B > T***

¹Human leukocyte antigen (HLA) genes not included, ²Alternative candidate autoimmunity gene in the region reported in the GWAS catalog or functional studies, ³Cis-regulatory SNPs with significant association with allele-specific gene expression in B or T cells, ⁴The random forest generates SNP importance scores based on the importance of each SNP for the prediction. The SNP scores are summed up over a gene region to obtain the final gene importance score, ⁵SLE = systemic lupus erythematosus, RA = rheumatoid arthritis, IBD = inflammatory bowel disease, CD = Crohn’s disease, T1D = diabetes mellitus type 1, MS = multiple sclerosis, PBC = primary biliary cirrhosis, UC = ulcerative colitis, KD = Kawasaki disease, Celiac = Celiac disease, AS = Ankylosing spondylitis, pSS = primary Sjögren’s syndrome, New = previously unknown SLE risk gene, ⁶Langefeld, C. D. et al. Transancestral mapping and genetic load in systemic lupus erythematosus, submitted manuscript, ⁷Evidence of SLE association from literature⁴⁴, ⁸Genes are annotated according to their expression level in B or T cells based on RNA-sequencing data, ⁹Low = Expression below 1 fragments per kilobase of exon per million fragments mapped (FPKM) for both cell types, ^*Bonferroni corrected p-value < 0.05, ^***Bonferroni corrected p-value < 0.001.

Back to article page

Table 1 Top 40 risk genes for SLE identified by random forest prediction using Immunochip genotype data from SLE patients and controls.

Search

Quick links