Transcriptome-Wide Association Study Reveals New Molecular Interactions Associated with Melanoma Pathogenesis

A transcriptome-wide association study (TWAS) was conducted on genome-wide association study (GWAS) summary statistics of malignant melanoma of skin (UK Biobank dataset) and The Cancer Genome Atlas-Skin Cutaneous Melanoma (TCGA-SKCM) gene expression weights to identify melanoma susceptibility genes. The GWAS included 2465 cases and 449,799 controls, while the gene expression testing was conducted on 103 cases. Afterward, a gene enrichment analysis was applied to identify significant TWAS associations. The melanoma's gene-microRNA (miRNA) regulatory network was constructed from the TWAS genes and their corresponding miRNAs. At last, a disease enrichment analysis was conducted on the corresponding miRNAs. The TWAS detected 27 genes associated with melanoma with p-values less than 0.05 (the top three genes are LOC389458 (RBAK), C16orf73 (MEIOB), and EIF3CL). After the joint/conditional test, one gene (AMIGO1) was dropped, resulting in 26 significant genes. The Gene Ontology (GO) biological process associated the extended gene set (76 genes) with protein K11-linked ubiquitination and regulation of cell cycle phase transition. K11-linked ubiquitin chains regulate cell division. Interestingly, the extended gene set was related to different skin cancer subtypes. Moreover, the enriched pathways were nsp1 from SARS-CoV-2 that inhibit translation initiation in the host cell, cell cycle, translation factors, and DNA repair pathways full network. The gene-miRNA regulatory network identified 10 hotspot genes with the top three: TP53, BRCA1, and MDM2; and four hotspot miRNAs: mir-16, mir-15a, mir-125b, and mir-146a. Melanoma was among the top ten diseases associated with the corresponding (106) miRNAs. Our results shed light on melanoma pathogenesis and biologically significant molecular interactions.

Methods for Assessing Population Relationships and History Using Genomic Data

Genetic data contain a record of our evolutionary history. The availability of large-scale datasets of human populations from various geographic areas and timescales, coupled with advances in the computational methods to analyze these data, has transformed our ability to use genetic data to learn about our evolutionary past. Here, we review some of the widely used statistical methods to explore and characterize population relationships and history using genomic data. We describe the intuition behind commonly used approaches, their interpretation, and important limitations. For illustration, we apply some of these techniques to genome-wide autosomal data from 929 individuals representing 53 worldwide populations that are part of the Human Genome Diversity Project. Finally, we discuss the new frontiers in genomic methods to learn about population history. In sum, this review highlights the power (and limitations) of DNA to infer features of human evolutionary history, complementing the knowledge gleaned from other disciplines, such as archaeology, anthropology, and linguistics.

Statistical Learning Methods for Neuroimaging Data Analysis with Applications

The aim of this review is to provide a comprehensive survey of statistical challenges in neuroimaging data analysis, from neuroimaging techniques to large-scale neuroimaging studies and statistical learning methods. We briefly review eight popular neuroimaging techniques and their potential applications in neuroscience research and clinical translation. We delineate four themes of neuroimaging data and review major image processing analysis methods for processing neuroimaging data at the individual level. We briefly review four large-scale neuroimaging-related studies and a consortium on imaging genomics and discuss four themes of neuroimaging data analysis at the population level. We review nine major population-based statistical analysis methods and their associated statistical challenges and present recent progress in statistical methodology to address these challenges.

Sparse Regression in Cancer Genomics: Comparing Variable Selection and Predictions in Real World Data

BACKGROUND

Evaluation of gene interaction models in cancer genomics is challenging, as the true distribution is uncertain. Previous analyses have benchmarked models using synthetic data or databases of experimentally verified interactions - approaches which are susceptible to misrepresentation and incompleteness, respectively. The objectives of this analysis are to (1) provide a real-world data-driven approach for comparing performance of genomic model inference algorithms, (2) compare the performance of LASSO, elastic net, best-subset selection,L 0 L 1 penalisation andL 0 L 2 penalisation in real genomic data and (3) compare algorithmic preselection according to performance in our benchmark datasets to algorithmic selection by internal cross-validation.

METHODS

Five large ( n 4000 ) genomic datasets were extracted from Gene Expression Omnibus. 'Gold-standard' regression models were trained on subspaces of these datasets ( n 4000 , p = 500 ). Penalised regression models were trained on small samples from these subspaces ( n ∈ { 25 , 75 , 150 } , p = 500 ) and validated against the gold-standard models. Variable selection performance and out-of-sample prediction were assessed. Penalty 'preselection' according to test performance in the other 4 datasets was compared to selection internal cross-validation error minimisation.

RESULTS

L 1 L 2 -penalisation achieved the highest cosine similarity between estimated coefficients and those of gold-standard models.L 0 L 2 -penalised models explained the greatest proportion of variance in test responses, though performance was unreliable in low signal:noise conditions.L 0 L 2 also attained the highest overall median variable selection F1 score. Penalty preselection significantly outperformed selection by internal cross-validation in each of 3 examined metrics.

CONCLUSIONS

This analysis explores a novel approach for comparisons of model selection approaches in real genomic data from 5 cancers. Our benchmarking datasets have been made publicly available for use in future research. Our findings support the use ofL 0 L 2 penalisation for structural selection andL 1 L 2 penalisation for coefficient recovery in genomic data. Evaluation of learning algorithms according to observed test performance in external genomic datasets yields valuable insights into actual test performance, providing a data-driven complement to internal cross-validation in genomic regression tasks.