Figure 1.
Schematic outline of the analysis pipeline.
The summary of the steps involved in this study. All human lncRNAs (sense exonic, sense non-exonic, antisense and intergenic) located within and in close proximity (5 kb up/downstream) of the IBD and T1D candidate genes were identified. Based on the above described workflow, we predicted potential lncRNA secondary structure-disruptive GWAS SNPs within the IBD and T1D loci-associated lncRNAs. cis-eQTL signals were identified for these lncRNAs and linkage disequilibrium analysis was also explored for selected SNPs. We exploited HapMap data to identify candidate lncRNAs under recent positive selection.
Figure 2.
Mapped lncRNA intervals within IBD and T1D susceptibility loci genes.
Categorization of the genomic association of the lncRNAs to the IBD and T1D loci-associated genes, as sense exonic (A), sense non-exonic (B), antisense (C), and intergenic (D).
Table 1.
IBD and T1D candidate gene associated lncRNAs.
Figure 3.
Structure disruption of lncRNA NONHSAG044354-associated with BACH2 (implicated in both IBD and T1D) by GWAS SNPs rs3757247 and rs597325.
The structure-disruptive effects of SNPs rs3757247 (A) and rs597325 (B) located in lncRNA NONHSAG044354 associated with candidate gene BACH2. (a) UCSC genome browser view showing the location of the predicted local region disrupted by the SNP. The color of predicted local region (green in this case) is based on the RNAsnp p-value (0.129 and 0.120 for rs3757247 and rs597325 respectively). (b) Minimum free energy structures (MFE) of the global wild-type and mutant sequences displaying the secondary structure and the local region affected by the SNP position, colored green (wild-type) and red (mutant) (c) Dot plot representing the base pair probabilities of wild type and mutant RNA sequences corresponding to the predicted local region by RNAsnp. The upper and lower triangle of the matrix represents the base pair probabilities of wild-type (green) and mutant (red), respectively.
Figure 4.
Regional LD plot for SNP rs3757247 associated with IBD and T1D loci.
The regional LD plot for SNP rs3757247 was calculated using HapMap3 (release 2) in SNAP tool. SNPs are represented as diamonds and the brightness of each SNP is proportional to the r2 threshold value for that SNP. SNPs rs11755527 and rs1847472 had an r2 value of 0.949 (T1D p-valuemeta 5.38e-08, Barrett et al 2009 [52]) and 0.565 (IBD p-valueichip 1.19e-04, IBD p-valuemeta 1.10e-09, Jostins et al 2012 [71]) respectively (shown in red).
Table 2.
Common structure-disruptive SNPs within IBD and T1D loci-associated lncRNAs.
Figure 5.
cis-eQTLs and gene-SNP associations for rs3757247 and rs597325 with BACH2 candidate gene.
(A) The gene SNP association for rs3757247 and BACH2 candidate gene in whole blood. (B, C, D) The gene SNP association for rs597325 and BACH2 in colon transverse, brain cortex and cell line fibroblasts. The gene SNP associations are calculated using linear regression model using Genotype Tissue Expression Portal (GTEx) in the selected tissues with more than 80 samples using a cis window of +/−1 MB around the transcription start site (TSS) at significance level of 0.05. For each gene SNP association plot, p-values are displayed at the bottom.
Table 3.
Structure-disruptive SNPs and their-associated lncRNAs.
Figure 6.
Expression of antisense IBD and T1D loci-associated lncRNAs.
Expression profiles for (A) 49 antisense IBD loci-associated lncRNAs and (B) 37 antisense T1D loci-associated lncRNAs expressed across all the HBM tissues. The FPKM threshold of >1 was used and the values were log10 transformed.
Figure 7.
Correlations of expression for antisense IBD and T1D loci-associated lncRNAs.
Spearman correlations for (A) antisense IBD and (B) antisense T1D loci-associated lncRNAs expressed across all the HBM tissues. The FPKM threshold of >1 was used and the values were log10 transformed.
Figure 8.
Comparison of expression levels of NONHSAG044354 with its host gene (BACH2) in 14 tissues.
(A)Tissue-specific gene expression profile of BACH2 candidate gene and NONHSAG044354 associated lncRNA across 14 tissues based on Human Body Map (HBM) data. (B) Sense exonic lncRNA NONHSAG044354 positively correlating (rs = 0.85) with BACH2 candidate gene. Protein coding mRNA expression is plotted on the x-axis and lncRNA expression is shown on the y-axis both on log10 FPKM for 14 HBM tissues.