Background Psoriasis (PsO): is an immune-mediated inflammatory disease that imposes a significant burden on patients. Many patients experience relapse or inadequate responses, and PsO subtypes also lack effective therapies, highlighting the need for new therapeutic targets. Methods: We performed a proteome-wide Mendelian randomization (MR) to explore potential therapeutic targets for PsO. Protein quantitative trait loci (pQTLs) data were obtained from the Pharma Proteomics Project (54,219 UK Biobank participants, 2923 proteins), and PsO phenotype and subtype data were sourced from FinnGen (10,312 cases; 397,564 controls) for discovery. Replication MR utilized integrated protein data (Iceland and Norfolk) and phenotype data from multiple databases (UK Biobank and GWAS Catalog). Reverse MR and colocalization were used to support causal relationships. Single-cell RNA-seq analysis revealed distinct expression patterns of protein-coding genes across different cell types in PsO biopsy samples and normal skin tissues. Protein-protein interactions (PPI) and molecular docking were used to evaluate druggability. Results: MR analysis identified 13 proteins significantly associated with PsO risk (p < $2.56\times {10}^{-5}$), including 10 proteins associated with PsO subtypes. Decreased levels of eight proteins (IFNLR1, APOF, TDRKH, DDR1, HLA-E, LTA, MOG, and ICAM3) and increased levels of five proteins (IFNGR2, HCG22, IL12B, BTN3A2, and TRIM40) showed protective effects against PsO progression. Robust colocalization ($PP{H}_4$ > 0.9) identified IFNLR1, IFNGR2, APOF, and TDRKH as top candidates. Single-cell RNA sequencing analysis revealed that IFNLR1, IFNGR2, LTA, TDRKH, and DDR1 were specifically expressed in T cells of psoriatic biopsy specimens compared to healthy controls. Molecular docking indicated the druggability of IFNLR1 and IFNGR2. Conclusions: We identified several potential therapeutic targets for PsO, with IFNLR1, IFNGR2, APOF, and TDRKH emerging as promising candidates, particularly IFNLR1 and IFNGR2, which are associated with the IFN family. These findings may provide new perspectives on PsO therapy and pathogenesis.

Coronary atherosclerosis (CAS) is a major cause of cardiovascular morbidity worldwide. The understanding of atherosclerosis has shifted from a cholesterol deposition disorder to an inflammation-driven disease, with anti-inflammatory therapies demonstrating clinical efficacy. Identifying inflammatory protein targets is crucial for developing targeted therapies. A proteome-wide Mendelian randomization (MR) analysis was performed to explore therapeutic targets for CAS by integrating inflammatory proteomics data from the UK-PPP (54,219 participants, 2923 proteins) and Iceland cohorts (35,559 participants, 4907 proteins) as exposures and outcome data for CAS, atherosclerosis, and carotid atherosclerosis from FinnGen. Replication MR employed meta-analysis of six proteomics datasets and CAS data from three sources, while the impact of the identified proteins on four cardiovascular diseases was also investigated. Colocalization analysis (PPH₄ > 0.9), reverse MR, and SMR were used to ensure robust causal inference. Proteome-wide MR identified 11 proteins significantly associated with CAS (p < 3.52 × 10⁻⁵), with all but CD4 linked to cardiovascular disease risk. Notably, colocalization confirmed the causal roles of PCSK9, IL6R, CELSR2, FN1, and SPARCL1 in CAS, and single-cell RNA-seq analysis revealed that five genes (TGFB1, SPARCL1, IL6R, FN1, and CELSR2) were exclusively expressed in smooth muscle cells of either coronary plaques or healthy vasculature. Druggability assessments were subsequently conducted for these targets. The three most promising targets (CELSR2, FN1, and SPARCL1), along with the other identified proteins and their biological functions, exhibit robust causal associations with CAS. FN1 and TGFB1 have the potential for drug repurposing in atherosclerosis treatment.

Bulk RNA-sequencing (bulk RNA-seq) averages gene expression across cell mixtures, obscuring single-cell heterogeneity and spatial architectures essential for understanding pathological processes. We developed HistoMap, a deep learning-based framework for single-cell spatial deconvolution. The model employs a two-stage pipeline: first, reconstructing high-fidelity single-cell profiles from bulk data using a β-variational autoencoder, and second, utilizing a Histological Vision Transformer (H-ViT) to map these cells to tissue coordinates via dual guidance from transcriptomic references and H&E-stained morphological constraints. HistoMap demonstrated superior performance across diverse human tissues, achieving a Pearson Correlation Coefficient (PCC) of 0.800 on external validation. Application to 14 colorectal cancer cases revealed a Macro_SPP1-mediated desmoplastic barrier. SPP1+ macrophages act as spatial hubs at the invasive front, forming a physical “sequestration belt” that functionally excludes cytotoxic T cells from the tumor core. HistoMap successfully bridges bulk RNA-seq and spatial single-cell architectures. Our findings provide a molecular rationale for immune checkpoint blockade resistance and identify the SPP1-fibroblast axis as a pivotal target for therapeutic sensitization.