Identification and verification of prognostic cancer subtype based on multi-omics analysis for kidney renal papillary cell carcinoma

On integrative analysis of multi-level gene expression data in Kidney cancer subgrouping

Kidney cancer is one of the most dangerous cancer mainly targeting men. In 2020, around 430, 000 people were diagnosed with this disease worldwide. It can be divided into three prime subgroups such as kidney renal cell carcinoma (KIRC), kidney renal papilliary cell carcinoma (KIRP) and kidney chromophobe (KICH). Correct identification of these subgroups on time is crucial for the initiation and determination of proper treatment. On-time identification of this disease and its subgroup can help both the clinicians and patients to improve the situation. Hence, this study checks the possibility of using multi-omics data in the kidney cancer subgrouping, whether integrating multiple omics data will increase the subgrouping accuracy or not. Four different molecular data such as genomics, proteomics, epigenomics and miRNA from The Cancer Genome Atlas (TCGA) are used in this study. As the data is in a very high dimension world, this study starts with selecting the relevant features of the study using Pearson's correlation coefficient. Those selected features are used with three different classification algorithms such as k-nearest neighbor (KNN), supporting vector machines (SVMs) and random forest. Performances are compared to see whether the integration of multi-omics data can improve the accuracy of kidney cancer subgrouping. This study shows that integration of multi-omics data can improve the performance of the kidney cancer subgrouping. The highest performance (accuracy value of 0.98±0.03) is gained by top 400 features selected from integrated multi-omics data, with support vector machines.

DeSide: A unified deep learning approach for cellular deconvolution of tumor microenvironment

Cellular deconvolution via bulk RNA sequencing (RNA-seq) presents a cost-effective and efficient alternative to experimental methods such as flow cytometry and single-cell RNA-seq (scRNA-seq) for analyzing the complex cellular composition of tumor microenvironments. Despite challenges due to heterogeneity within and among tumors, our innovative deep learning-based approach, DeSide, shows exceptional accuracy in estimating the proportions of 16 distinct cell types and subtypes within solid tumors. DeSide integrates biological pathways and assesses noncancerous cell types first, effectively sidestepping the issue of highly variable gene expression profiles (GEPs) associated with cancer cells. By leveraging scRNA-seq data from six cancer types and 185 cancer cell lines across 22 cancer types as references, our method introduces distinctive sampling and filtering techniques to generate a high-quality training set that closely replicates real tumor GEPs, based on The Cancer Genome Atlas (TCGA) bulk RNA-seq data. With this model and high-quality training set, DeSide outperforms existing methods in estimating tumor purity and the proportions of noncancerous cells within solid tumors. Our model precisely predicts cellular compositions across 19 cancer types from TCGA and proves its effectiveness with multiple additional external datasets. Crucially, DeSide enables the identification and analysis of combinatorial cell type pairs, facilitating the stratification of cancer patients into prognostically significant groups. This approach not only provides deeper insights into the dynamics of tumor biology but also highlights potential therapeutic targets by underscoring the importance of specific cell type or subtype interactions.