Pan-cancer analysis for studying cancer stage using protein and gene expression data

Multi-omics analysis reveals a molecular landscape of the early recurrence and early metastasis in pan-cancer

Cancer remains a formidable challenge in medicine due to its propensity for recurrence and metastasis, which can result in unfavorable treatment outcomes. This challenge is particularly acute for early-stage patients, who may experience recurrence and metastasis without timely detection. Here, we first analyzed the differences in clinical characteristics among the primary tumor, recurrent tumor, and metastatic tumor in different stages of cancer, which may be caused by the molecular level. Moreover, the importance of predicting early cancer recurrence and metastasis is emphasized by survival analyses. Next, we used a multi-omics approach to identify key molecular changes associated with early cancer recurrence and metastasis and discovered that early metastasis in cancer demonstrated a high degree of genomic and cellular heterogeneity. We performed statistical comparisons for each level of omics data including gene expression, mutation, copy number variation, immune cell infiltration, and cell status. Then, various analytical techniques, such as proportional hazard model and Fisher's exact test, were used to identify specific genes or immune characteristics associated with early cancer recurrence and metastasis. For example, we observed that the overexpression of BPIFB1 and high initial B-cell infiltration levels are linked to early cancer recurrence, while the overexpression or amplification of ANKRD22 and LIPM, mutation of IGHA1 and MUC16, high fibroblast infiltration level, M1 polarization of macrophages, cellular status of DNA repair are all linked to early cancer metastasis. These findings have led us to construct classifiers, and the average area under the curve (AUC) of these classifiers was greater than 0.75 in The Cancer Genome Atlas (TCGA) cancer patients, confirming that the features we identified could be biomarkers for predicting recurrence and metastasis of early cancer. Finally, we identified specific early sensitive targets for targeted therapy and immune checkpoint inhibitor therapy. Once the biomarkers we identified changed, treatment-sensitive targets can be treated accordingly. Our study has comprehensively characterized the multi-omics characteristics and identified a panel of biomarkers of early cancer recurrence and metastasis. Overall, it provides a valuable resource for cancer recurrence and metastasis research and improves our understanding of the underlying mechanisms driving early cancer recurrence and metastasis.

Joint learning sample similarity and correlation representation for cancer survival prediction

BACKGROUND

As a highly aggressive disease, cancer has been becoming the leading death cause around the world. Accurate prediction of the survival expectancy for cancer patients is significant, which can help clinicians make appropriate therapeutic schemes. With the high-throughput sequencing technology becoming more and more cost-effective, integrating multi-type genome-wide data has been a promising method in cancer survival prediction. Based on these genomic data, some data-integration methods for cancer survival prediction have been proposed. However, existing methods fail to simultaneously utilize feature information and structure information of multi-type genome-wide data.

RESULTS

We propose a Multi-type Data Joint Learning (MDJL) approach based on multi-type genome-wide data, which comprehensively exploits feature information and structure information. Specifically, MDJL exploits correlation representations between any two data types by cross-correlation calculation for learning discriminant features. Moreover, based on the learned multiple correlation representations, MDJL constructs sample similarity matrices for capturing global and local structures across different data types. With the learned discriminant representation matrix and fused similarity matrix, MDJL constructs graph convolutional network with Cox loss for survival prediction.

CONCLUSIONS

Experimental results demonstrate that our approach substantially outperforms established integrative methods and is effective for cancer survival prediction.

Deciphering molecular mechanisms of metastasis: novel insights into targets and therapeutics

BACKGROUND

The transition of a primary tumour to metastatic progression is driven by dynamic molecular changes, including genetic and epigenetic alterations. The metastatic cascade involves bidirectional interactions among extracellular and intracellular components leading to disintegration of cellular junctions, cytoskeleton reorganization and epithelial to mesenchymal transition. These events promote metastasis by reprogramming the primary cancer cell's molecular framework, enabling them to cause local invasion, anchorage-independent survival, cell death and immune resistance, extravasation and colonization of distant organs. Metastasis follows a site-specific pattern that is still poorly understood at the molecular level. Although various drugs have been tested clinically across different metastatic cancer types, it has remained difficult to develop efficacious therapeutics due to complex molecular layers involved in metastasis as well as experimental limitations.

CONCLUSIONS

In this review, a systemic evaluation of the molecular mechanisms of metastasis is outlined and the potential molecular components and their status as therapeutic targets and the associated pre-clinical and clinical agents available or under investigations are discussed. Integrative methods like pan-cancer data analysis, which can provide clinical insights into both targets and treatment decisions and help in the identification of crucial components driving metastasis such as mutational profiles, gene signatures, associated pathways, site specificities and disease-gene phenotypes, are discussed. A multi-level data integration of the metastasis signatures across multiple primary and metastatic cancer types may facilitate the development of precision medicine and open up new opportunities for future therapies.

Deep learning based feature-level integration of multi-omics data for breast cancer patients survival analysis

BACKGROUND

Breast cancer is the most prevalent and among the most deadly cancers in females. Patients with breast cancer have highly variable survival lengths, indicating a need to identify prognostic biomarkers for personalized diagnosis and treatment. With the development of new technologies such as next-generation sequencing, multi-omics information are becoming available for a more thorough evaluation of a patient's condition. In this study, we aim to improve breast cancer overall survival prediction by integrating multi-omics data (e.g., gene expression, DNA methylation, miRNA expression, and copy number variations (CNVs)).

METHODS

Motivated by multi-view learning, we propose a novel strategy to integrate multi-omics data for breast cancer survival prediction by applying complementary and consensus principles. The complementary principle assumes each -omics data contains modality-unique information. To preserve such information, we develop a concatenation autoencoder (ConcatAE) that concatenates the hidden features learned from each modality for integration. The consensus principle assumes that the disagreements among modalities upper bound the model errors. To get rid of the noises or discrepancies among modalities, we develop a cross-modality autoencoder (CrossAE) to maximize the agreement among modalities to achieve a modality-invariant representation. We first validate the effectiveness of our proposed models on the MNIST simulated data. We then apply these models to the TCCA breast cancer multi-omics data for overall survival prediction.

RESULTS

For breast cancer overall survival prediction, the integration of DNA methylation and miRNA expression achieves the best overall performance of 0.641 ± 0.031 with ConcatAE, and 0.63 ± 0.081 with CrossAE. Both strategies outperform baseline single-modality models using only DNA methylation (0.583 ± 0.058) or miRNA expression (0.616 ± 0.057).

CONCLUSIONS

In conclusion, we achieve improved overall survival prediction performance by utilizing either the complementary or consensus information among multi-omics data. The proposed ConcatAE and CrossAE models can inspire future deep representation-based multi-omics integration techniques. We believe these novel multi-omics integration models can benefit the personalized diagnosis and treatment of breast cancer patients.

Integrating multi-omics data by learning modality invariant representations for improved prediction of overall survival of cancer

Breast and ovarian cancers are the second and the fifth leading causes of cancer death among women. Predicting the overall survival of breast and ovarian cancer patients can facilitate the therapeutics evaluation and treatment decision making. Multi-scale multi-omics data such as gene expression, DNA methylation, miRNA expression, and copy number variations can provide insights on personalized survival. However, how to effectively integrate multi-omics data remains a challenging task. In this paper, we develop multi-omics integration methods to improve the prediction of overall survival for breast cancer and ovarian cancer patients. Because multi-omics data for the same patient jointly impact the survival of cancer patients, features from different -omics modality are related and can be modeled by either association or causal relationship (e.g., pathways). By extracting these relationships among modalities, we can get rid of the irrelevant information from high-throughput multi-omics data. However, it is infeasible to use the Brute Force method to capture all possible multi-omics interactions. Thus, we use deep neural networks with novel divergence-based consensus regularization to capture multi-omics interactions implicitly by extracting modality-invariant representations. In comparing the concatenation-based integration networks with our new divergence-based consensus networks, the breast cancer overall survival C-index is improved from 0.655±0.062 to 0.671±0.046 when combing DNA methylation and miRNA expression, and from 0.627±0.062 to 0.667±0.073 when combing miRNA expression and copy number variations. In summary, our novel deep consensus neural network has successfully improved the prediction of overall survival for breast cancer and ovarian cancer patients by implicitly learning the multi-omics interactions.

An Integrated Pan-Cancer Analysis and Structure-Based Virtual Screening of GPR15

G protein-coupled receptor 15 (GPR15, also known as BOB) is an extensively studied orphan G protein-coupled receptors (GPCRs) involving human immunodeficiency virus (HIV) infection, colonic inflammation, and smoking-related diseases. Recently, GPR15 was deorphanized and its corresponding natural ligand demonstrated an ability to inhibit cancer cell growth. However, no study reported the potential role of GPR15 in a pan-cancer manner. Using large-scale publicly available data from the Cancer Genome Atlas (TCGA) and the Genotype-Tissue Expression (GTEx) databases, we found that GPR15 expression is significantly lower in colon adenocarcinoma (COAD) and rectal adenocarcinoma (READ) than in normal tissues. Among 33 cancer types, GPR15 expression was significantly positively correlated with the prognoses of COAD, neck squamous carcinoma (HNSC), and lung adenocarcinoma (LUAD) and significantly negatively correlated with stomach adenocarcinoma (STAD). This study also revealed that commonly upregulated gene sets in the high GPR15 expression group (stratified via median) of COAD, HNSC, LUAD, and STAD are enriched in immune systems, indicating that GPR15 might be considered as a potential target for cancer immunotherapy. Furthermore, we modelled the 3D structure of GPR15 and conducted structure-based virtual screening. The top eight hit compounds were screened and then subjected to molecular dynamics (MD) simulation for stability analysis. Our study provides novel insights into the role of GPR15 in a pan-cancer manner and discovered a potential hit compound for GPR15 antagonists.

Failed immune responses across multiple pathologies share pan-tumor and circulating lymphocytic targets

Rationale Tumor infiltrating lymphocytes are widely associated with positive outcomes, yet carry key indicators of a systemic failed immune response against unresolved cancer. Cancer immunotherapies can reverse their tolerance phenotypes, while preserving tumor-reactivity and neoantigen-specificity shared with circulating immune cells. Objectives We performed comprehensive transcriptomic analyses to identify gene signatures common to circulating and tumor infiltrating lymphocytes in the context of clear cell renal cell carcinoma. Modulated genes also associated with disease outcome were validated in other cancer types. Findings Using bioinformatics, we identified practical diagnostic markers and actionable targets of the failed immune response. On circulating lymphocytes, three genes, LEF1, FASLG, and MMP9, could efficiently stratify patients from healthy control donors. From their associations with resistance to cancer immunotherapies and microbial infections, we uncovered not only pan-cancer, but pan-pathology failed immune response profiles. A prominent lymphocytic matrix metallopeptidase cell migration pathway, is central to a panoply of diseases and tumor immunogenicity, correlates with multi-cancer recurrence, and identifies a feasible, non-invasive approach to pan-pathology diagnoses. Conclusions The non-invasive differently expressed genes we have identified warrant future investigation towards the development of their potential in precision diagnostics and precision pan-disease immunotherapeutics.