BGPT: Paper Review: Sequencing-based approach identified three new susceptibility loci for psoriasis

Explore by Goal

Quick Explanation Copied

Key result

This multi-stage sequencing study in Chinese cohorts reports three new psoriasis susceptibility loci—NFKB1 (4q24) at rs1020760, a CD27–LAG3 region (12p13.3) at rs758739, and IKZF3 (17q12) at rs10852936—while also confirming known loci (IL12B, IFIH1, ERAP1, RNF114).

However, the paper does not provide direct functional proof of causality for the specific noncoding variants; it mainly connects association signals to candidate genes via in-silico/annotation evidence.

See: 10.1038/ncomms5331

Long Explanation

Sequencing-based approach identified three new susceptibility loci for psoriasis

Nature Communications (2014-07-09). DOI: 10.1038/ncomms5331

What the authors claim: three additional psoriasis susceptibility loci in Han Chinese cohorts, plus replication of previously known loci.

1) Evidence snapshot (association strength)

From the provided full text tables: combined p-values and ORs for the main reported loci.

Directionality is modest: ORs are close to 1 (e.g., NFKB1 rs1020760 OR ≈ 1.12; IKZF3 rs10852936 OR ≈ 1.10).
Strength differs by locus: the known IL12B locus shows extremely strong evidence in the paper (p ~ 10^-19), providing an internal benchmark that the study can detect large signals.

2) Study design, population scope, and statistical logic

What was sequenced, how candidates were selected, and what was replicated.

Cohorts (as reported)

Stage	Cases	Controls	Purpose
Exome sequencing discovery	781	676	Initial discovery of noncoding variants in targeted regions
Targeted sequencing	9,946	9,906	Increase power and coverage in prioritized gene-centric regions
Genotyping replication	4,480	6,521	Replication/validation of top candidate SNVs

Key methodological choice: rather than scanning the entire genome for all noncoding variation, the study focuses on noncoding SNVs in/near target genes and uses a staged design to prioritize and replicate.

Why that matters for interpretation: association signals can point to regulatory elements, but “near-gene” choice can bias which biology is discoverable (and can miss distal enhancers). This is an important blind spot when trying to move from “locus” to “mechanism.”

3) From locus → candidate biology (what the paper actually provides)

Reported candidate-gene narratives use ENCODE regulatory annotations and public psoriasis expression datasets; causality remains unproven.

NFKB1 (rs1020760, 4q24)

Paper argues NF-κB involvement because NFKB1 encodes components of NF-κB and because NF-κB regulates pro-inflammatory gene expression relevant to psoriasis, with additional expression/ENCODE-based support.

CD27–LAG3 region (rs758739, 12p13.3)

Paper highlights CD27 and LAG3 as plausible immune regulators (co-stimulation vs inhibitory regulation), supported by expression differences in psoriasis and ENCODE DHS locality for one lead SNP.

IKZF3 (rs10852936, 17q12)

Paper proposes IKZF3 as candidate based on its known biology in immune cell regulation and mentions psoriasis-associated expression differences and ENCODE DHS/enhancer signal context.

4) Skeptical critique: what’s strong vs what remains uncertain

Strengths

Replication in an independent Han Chinese cohort supports robustness of the reported lead signals.
Internal calibration: known loci (IL12B, IFIH1, ERAP1, RNF114) are re-detected strongly, suggesting the pipeline can capture real signals.

Uncertainties / potential failure modes

Causal variant problem: the paper reports associations for specific noncoding SNVs but does not provide direct functional experiments showing that those variants change gene regulation in disease-relevant cells.
Targeted-region ascertainment bias: the analysis is “gene-centric” (flanking regions within 200 bp; targeted sequencing of 1,362 genes). That can reduce false positives relative to whole-genome scans, but it also constrains discovery to what was selected/covered.
Population scope: replication is within Han Chinese cohorts. Genetic architecture and LD differ across ancestries; thus, transferability to other populations is not established by this paper alone.
Mechanistic inference is annotation-dependent: ENCODE DHS/enhancer signals and public expression comparisons are suggestive, but they can be cell-line- or context-dependent and may not reflect the exact causal tissue state in psoriasis lesions.

This paper adds three loci, but later work shows that psoriasis genetics comprises many loci genome-wide. One recent example is a large GWAS meta-analysis reporting 109 distinct susceptibility loci (mostly European ancestry).

5) “Locus-to-function” in psoriasis: what later evidence suggests

The paper predates many allele-to-gene mapping methods. Later studies illustrate what kind of follow-up is needed.

Example of functional follow-up (DNA interaction mapping)

A later psoriasis study (Capture Hi-C + CRISPR activation + expression analysis) mapped a noncoding susceptibility locus to a target gene (KLF4) using DNA interaction landscapes and functional perturbation in relevant immune/skin cell models.

Implication for the 2014 paper: similar allele-specific validation (e.g., regulatory element editing/activation tied to NF-κB, CD27/LAG3, and IKZF3 in psoriasis-relevant cell states) would be needed to elevate these loci from association to causality.

Paper review metrics

Novelty (1–10): 8
Explanation: adds three new loci in a staged sequencing/replication framework for noncoding SNVs, expanding the early psoriasis locus catalog with an East-Asian genetic focus.

Scientific quality (1–10): 8
Explanation: large multi-stage sample sizes with replication, good internal validation via known loci, and use of LD/conditional checks; main weakness is lack of direct functional causality experiments for the lead noncoding variants.

Generality (1–10): 7
Explanation: locus discoveries can generalize via shared biology (NF-κB/immune regulation) but replication is within Han Chinese cohorts; generalizability across ancestries needs further replication.

Usefulness (1–10): 8
Explanation: provides concrete, testable loci (variants + LD blocks) and candidate pathways to prioritize future fine-mapping and functional studies.

Reproducibility (1–10): 8
Explanation: multi-stage design with explicit cohort sizes, reported QC metrics, accession to raw sequencing data (SRA), and named tools; remaining reproducibility limit is whether supplementary details cover every analytical parameter.

Explanatory depth (1–10): 7
Explanation: integrates association with plausible immune biology narratives (NF-κB; T-cell regulation; IKZF3) but does not experimentally establish the variant→gene regulation→cell phenotype chain.

Actionable follow-ups (what would disprove/advance this paper)

Disprove causality locally: if allele-specific perturbation of the lead variants (or the entire credible set refined by fine-mapping) fails to alter regulatory activity and/or relevant gene expression in psoriasis-relevant cell states, then “locus → mechanism” weakens.

Test portability: if replication in non-Han ancestries consistently fails to show association (or shows different lead variants due to LD differences), then locus effects may be ancestry-specific or require re-parameterization of credible sets.

Resolve target genes: because the paper’s candidate gene mapping is partly based on proximity and regulatory annotations, experimental dissection of which gene(s) are directly regulated by these loci is needed (analogous to later DNA interaction mapping approaches in psoriasis).

Author reviews (tap to open BGPT author-focused analysis)

Feedback:

Updated: March 28, 2026