Dynamic changes of the immune microenvironment in the development of gastric cancer caused by inflammation.

Precancerous lesions typically precede gastric cancer (GC), but the molecular mechanisms underlying the transition from these lesions to GC remain unclear. Therefore, it is urgent to understand this transition from precancerous lesions to GC, which is crucial for the early diagnosis and treatment of GC. In this study, we merged multiple single-cell RNA sequencing datasets to investigate the molecular changes in distinct cell types associated with the progression of GC. First, we observed an increasing abundance of immune cells and a decrease in non-immune cells from non-atrophic gastritis to GC. Five immune cell types were significantly enriched in GC compared to precancerous lesions. Moreover, we found that the interleukin (IL)-17 signaling pathway and Th17 cell differentiation were significantly up-regulated in immune cell subsets during GC progression. Some genes in these processes were predominantly expressed at the GC stage, highlighting their potential as diagnostic markers. Furthermore, we validated our findings using bulk RNA sequencing data from The Cancer Genome Atlas and confirmed consistent immune cell changes during GC progression. Our study provides insights into the immune infiltration and signaling pathways involved in the development of GC, contributing to the development of early diagnosis and targeted treatment strategies for this malignancy.

Pathological image profiling identifies onco-microbial, tumor immune microenvironment, and prognostic subtypes of colorectal cancer.

Histology slide, tissue microbes, and the host gene expression can be independent prognostic factors of colorectal cancer (CRC), but the underlying associations and biological significance of these multimodal omics remain unknown. Here, we comprehensively profiled the matched pathological images, intratumoral microbes, and host gene expression characteristics in 527 patients with CRC. By clustering these patients based on histology slide features, we classified the patients into two histology slide subtypes (HSS). Onco-microbial community and tumor immune microenvironment (TIME) were also significantly different between the two subtypes (HSS1 and HSS2) of patients. Furthermore, variation in intratumoral microbes-host interaction was associated with the prognostic heterogeneity between HSS1 and HSS2. This study proposes a new CRC classification based on pathological image features and elucidates the process by which tumor microbes-host interactions are reflected in pathological images through the TIME.

Predictable regulation of survival by intratumoral microbe-immune crosstalk in patients with lung adenocarcinoma.

Intratumoral microbiota can regulate the tumor immune microenvironment (TIME) and mediate tumor prognosis by promoting inflammatory response or inhibiting anti-tumor effects. Recent studies have elucidated the potential role of local tumor microbiota in the development and progression of lung adenocarcinoma (LUAD). However, whether intratumoral microbes are involved in the TIME that mediates the prognosis of LUAD remains unknown. Here, we obtained the matched tumor microbiome and host transcriptome and survival data of 478 patients with LUAD in The Cancer Genome Atlas (TCGA). Machine learning models based on immune cell marker genes can predict 1- to 5-year survival with relative accuracy. Patients were stratified into high- and low-survival-risk groups based on immune cell marker genes, with significant differences in intratumoral microbial communities. Specifically, patients in the high-risk group had significantly higher alpha diversity (p < 0.05) and were characterized by an enrichment of lung cancer-related genera such as Streptococcus. However, network analysis highlighted a more active pattern of dominant bacteria and immune cell crosstalk in TIME in the low-risk group compared to the high-risk group. Our study demonstrated that intratumoral microbiota-immune crosstalk was strongly associated with prognosis in LUAD patients, which would provide new targets for the development of precise therapeutic strategies.

Single-cell RNA-seq reveals the metabolic status of immune cells response to immunotherapy in triple-negative breast cancer.

Immune checkpoint blockade (ICB) therapy offers promise in the treatment of triple-negative breast cancer (TNBC); however, its limited efficacy in certain TNBC patients poses a challenge. In this study, we elucidated the metabolic mechanism at 'sub-subtype' resolution underlying the non-response to ICB therapy in TNBC. Here, an analytic pipeline was developed to reveal the metabolic heterogeneity, which is correlated with the ICB outcomes, within each immune cell subtype. First, we identified metabolic 'sub-subtypes' within certain cell subtypes, predominantly T cell subsets, which are enriched in ICB non-responders and named as non-responder-enriched (NR-E) clusters. Notably, most of NR-E T metabolic cells exhibit globally higher metabolic activities compared to other cells within the same individual subtype. Further, we investigated the extra-cellular signals that trigger the metabolic status of NR-E T cells. In detail, the prediction of cell-to-cell communication indicated that NR-E T cells are regulated by plasmatic dendritic cells (pDCs) through TNFSF9, as well as by macrophages expressing SIGLEC9. In addition, we also validate the communication between TNFSF9+ pDCs and NR-E T cells utilizing deconvolution of spatial transcriptomics analysis. In summary, our research identified specific metabolic 'sub-subtypes' associated with ICB non-response and uncovered the mechanisms of their regulation in TNBC. And the proposed analytical pipeline can be used to examine metabolic heterogeneity within cell types that correlate with diverse phenotypes.

Enhancing Drug Repositioning through Local Interactive Learning with Bilinear Attention Networks.

Drug repositioning has emerged as a promising strategy for identifying new therapeutic applications for existing drugs. In this study, we present DRGBCN, a novel computational method that integrates heterogeneous information through a deep bilinear attention network to infer potential drugs for specific diseases. DRGBCN involves constructing a comprehensive drug-disease network by incorporating multiple similarity networks for drugs and diseases. Firstly, we introduce a layer attention mechanism to effectively learn the embeddings of graph convolutional layers from these networks. Subsequently, a bilinear attention network is constructed to capture pairwise local interactions between drugs and diseases. This combined approach enhances the accuracy and reliability of predictions. Finally, a multi-layer perceptron module is employed to evaluate potential drugs. Through extensive experiments on three publicly available datasets, DRGBCN demonstrates better performance over baseline methods in 10-fold cross-validation, achieving an average area under the receiver operating characteristic curve (AUROC) of 0.9399. Furthermore, case studies on bladder cancer and acute lymphoblastic leukemia confirm the practical application of DRGBCN in real-world drug repositioning scenarios. Importantly, our experimental results from the drug-disease network analysis reveal the successful clustering of similar drugs within the same community, providing valuable insights into drug-disease interactions. In conclusion, DRGBCN holds significant promise for uncovering new therapeutic applications of existing drugs, thereby contributing to the advancement of precision medicine.

Pan-cancer analyses reveal the stratification of patient prognosis by viral composition in tumor tissues.

The associations between cancer and bacteria/fungi have been extensively studied, but the implications of cancer-associated viruses have not been thoroughly examined. In this study, we comprehensively characterized the cancer virome of tissue samples across 31 cancer types, as well as blood samples from 23 cancer types. Our findings demonstrated the presence of viral DNA at low abundances in both tissue and blood across major human cancers, with significant differences in viral community composition observed among various cancer types. Furthermore, Cox regression analyses conducted on four cancers, including Head and Neck squamous cell carcinoma (HNSC), Kidney renal clear cell carcinoma (KIRC), Stomach adenocarcinoma (STAD), and Uterine Corpus Endometrial Carcinoma (UCEC), revealed strong correlation between viral composition/abundance in tissues and patient survival. Additionally, we identified virus-associated prognostic signatures (VAPS) for these four cancers, and discerned differences in the interplay between VAPS and dominant bacteria in tissues among patients with varying survival risks. Notably, clinically relevant analyses revealed prognostic capacities of the VAPS in these four cancers. Taken together, our study provides novel insights into the role of viruses in tissue in the prognosis of multiple cancers and offers guidance on the use of tissue viruses to stratify prognosis for patients with cancer.

Predicting gastric cancer tumor mutational burden from histopathological images using multimodal deep learning.

Tumor mutational burden (TMB) is a significant predictive biomarker for selecting patients that may benefit from immune checkpoint inhibitor therapy. Whole exome sequencing is a common method for measuring TMB; however, its clinical application is limited by the high cost and time-consuming wet-laboratory experiments and bioinformatics analysis. To address this challenge, we downloaded multimodal data of 326 gastric cancer patients from The Cancer Genome Atlas, including histopathological images, clinical data and various molecular data. Using these data, we conducted a comprehensive analysis to investigate the relationship between TMB, clinical factors, gene expression and image features extracted from hematoxylin and eosin images. We further explored the feasibility of predicting TMB levels, i.e. high and low TMB, by utilizing a residual network (Resnet)-based deep learning algorithm for histopathological image analysis. Moreover, we developed a multimodal fusion deep learning model that combines histopathological images with omics data to predict TMB levels. We evaluated the performance of our models against various state-of-the-art methods using different TMB thresholds and obtained promising results. Specifically, our histopathological image analysis model achieved an area under curve (AUC) of 0.749. Notably, the multimodal fusion model significantly outperformed the model that relied only on histopathological images, with the highest AUC of 0.971. Our findings suggest that histopathological images could be used with reasonable accuracy to predict TMB levels in gastric cancer patients, while multimodal deep learning could achieve even higher levels of accuracy. This study sheds new light on predicting TMB in gastric cancer patients.

A meta-analysis of tissue microbial biomarkers for recurrence and metastasis in multiple cancer types.

Background. Local recurrence and distant metastasis are the main causes of death in patients with cancer. Only considering species abundance changes when identifying markers of recurrence and metastasis in patients hinders finding solutions.Hypothesis. Consideration of microbial abundance changes and microbial interactions facilitates the identification of microbial markers of tumour recurrence and metastasis.Aim. This study aims to simultaneously consider microbial abundance changes and microbial interactions to identify microbial markers of recurrence and metastasis in multiple cancer types.Method. One thousand one hundred and six non-RM (patients without recurrence and metastasis within 3 years after initial surgery) tissue samples and 912 RM (patients with recurrence or metastasis within 3 years after initial surgery) tissue samples representing 11 cancer types were collected from The Cancer Genome Atlas (TCGA).Results. Tumour tissue bacterial composition differed significantly among 11 cancers. Among them, the tissue microbiome of four cancers, head and neck squamous cell carcinoma (HNSC), lung squamous cell carcinoma (LUSC), stomach adenocarcinoma (STAD) and uterine corpus endometrial carcinoma (UCEC), showed relatively good performance in predicting recurrence and metastasis in patients, with areas under the receiver operating characteristic curve (AUCs) of 0.78, 0.74, 0.91 and 0.93, respectively. Considering both species abundance changes and microbial interactions for the four cancers, a combination of nine genera (Niastella, Schlesneria, Thioalkalivibrio, Phaeobacter, Sphaerotilus, Thiomonas, Lawsonia, Actinobacillus and Spiroplasma) performed best in predicting patient survival.Conclusion. Taken together, our results imply that comprehensive consideration of microbial abundance changes and microbial interactions is helpful for mining bacterial markers that carry prognostic information.

Revealing the roles of TLR7, a nucleic acid sensor for COVID-19 in pan-cancer.

Recent studies suggested that cancer was a risk factor for coronavirus disease 2019 (COVID-19). Toll-like receptor 7 (TLR7), a severe acute respiratory syndrome 2 (SARS-CoV-2) virus's nucleic acid sensor, was discovered to be aberrantly expressed in many types of cancers. However, its expression pattern across cancers and association with COVID-19 (or its causing virus SARS-CoV-2) has not been systematically studied. In this study, we proposed a computational framework to comprehensively study the roles of TLR7 in COVID-19 and pan-cancers at genetic, gene expression, protein, epigenetic, and single-cell levels. We applied the computational framework in a few databases, including The Cancer Genome Atlas (TCGA), The Genotype-Tissue Expression (GTEx), Cancer Cell Line Encyclopedia (CCLE), Human Protein Atlas (HPA), lung gene expression data of mice infected with SARS-CoV-2, and the like. As a result, TLR7 expression was found to be higher in the lung of mice infected with SARS-CoV-2 than that in the control group. The analysis in the Opentargets database also confirmed the association between TLR7 and COVID-19. There are also a few exciting findings in cancers. First, the most common type of TLR7 was "Missense" at the genomic level. Second, TLR7 mRNA expression was significantly up-regulated in 6 cancer types and down-regulated in 6 cancer types compared to normal tissues, further validated in the HPA database at the protein level. The genes significantly co-expressed with TLR7 were mainly enriched in the toll-like receptor signaling pathway, endolysosome, and signaling pattern recognition receptor activity. In addition, the abnormal TLR7 expression was associated with mismatch repair (MMR), microsatellite instability (MSI), and tumor mutational burden (TMB) in various cancers. Mined by the ESTIMATE algorithm, the expression of TLR7 was also closely linked to various immune infiltration patterns in pan-cancer, and TLR7 was mainly enriched in macrophages, as revealed by single-cell RNA sequencing. Third, abnormal expression of TLR7 could predict the survival of Brain Lower Grade Glioma (LGG), Lung adenocarcinoma (LUAD), Skin Cutaneous Melanoma (SKCM), Stomach adenocarcinoma (STAD), and Testicular Germ Cell Tumors (TGCT) patients, respectively. Finally, TLR7 expressions were very sensitive to a few targeted drugs, such as Alectinib and Imiquimod. In conclusion, TLR7 might be essential in the pathogenesis of COVID-19 and cancers.

Application of oncolytic virus in tumor therapy.

Oncolytic viruses (OVs) can selectively kill tumor cells without affecting normal cells, as well as activate the innate and adaptive immune systems in patients. Thus, they have been considered as a promising measure for safe and effective cancer treatment. Recently, a few genetically engineered OVs have been developed to further improve the effect of tumor elimination by expressing specific immune regulatory factors and thus enhance the body's antitumor immunity. In addition, the combined therapies of OVs and other immunotherapies have been applied in clinical. Although there are many studies on this hot topic, a comprehensive review is missing on illustrating the mechanisms of tumor clearance by OVs and how to modify engineered OVs to further enhance their antitumor effects. In this study, we provided a review on the mechanisms of immune regulatory factors in OVs. In addition, we reviewed the combined therapies of OVs with other therapies including radiotherapy and CAR-T or TCR-T cell therapy. The review is useful in further generalize the usage of OV in cancer treatment.

MLSpatial: A machine-learning method to reconstruct the spatial distribution of cells from scRNA-seq by extracting spatial features.

MOTIVATION: Single-cell RNA sequencing (scRNA-seq) technologies allow us to interrogate the state of an individual cell within its microenvironment. However, prior to sequencing, cells should be dissociated first, making it difficult to obtain their spatial information. Since the spatial distribution of cells is critical in a few circumstances such as cancer immunotherapy, we present MLSpatial, a novel computational method to learn the relationship between gene expression patterns and spatial locations of cells, and then predict cell-to-cell distance distribution based on scRNA-seq data alone. RESULTS: We collected the drosophila embryo dataset, which contains both the fluorescence in situ hybridization (FISH) data and single cell RNA-seq (scRNA-seq) data of drosophila embryo. The FISH data provided the spatial position of 3039 cells and the expression of 84 genes for each cell. The scRNA-seq data contains the expressions of 8924 genes in 1297 high-quality cells with cell location unknown. For a comparison, we also collected the MERFISH data of 645 osteosarcoma cells with cell location and the expression status of 10,050 genes known. For each data, the cells were randomly divided into a training set and a test set, in the ratio of 7:3. The cell-to-cell distances our model extracted had a higher correspondence (i.e., correlation coefficient 0.99) with those of the real situation than those of existing methods in the FISH data of drosophila embryo. However, in the osteosarcoma data, our model captured the spatial relationship between cells, with a correlation of 0.514 to that of the real situation. We also applied the model trained using the FISH data of drosophila embryo into the single cell data of drosophila embryo, for which the real location of cells are unknown. The reconstructed pseudo drosophila embryo and the real embryo (as shown by the FISH data) had a high similarity in the spatial distribution of gene expression. CONCLUSION: MLSpatial can accurately restore the relative position of cells from scRNA-seq data; however, the performance depends on the type of cells. The trained model might be useful in reconstructing the spatial distributions of single cells with only scRNA-seq data, provided that the scRNA-seq data and the FISH data are under similar background (i.e., the same tissue with similar disease background).

Intratumoral Microbiota-Host Interactions Shape the Variability of Lung Adenocarcinoma and Lung Squamous Cell Carcinoma in Recurrence and Metastasis.

Differences in tissue microbiota-host interaction between lung squamous cell carcinoma (LUSC) and lung adenocarcinoma (LUAD) about recurrence and metastasis have not been well studied. In this study, we performed bioinformatics analyses to identify the genes and tissue microbes significantly associated with recurrence or metastasis. All lung cancer patients were divided into the recurrence or metastasis (RM) group and the nonrecurrence and nonmetastasis (non-RM) group according to whether or not they had recurred or metastasized within 3 years after the initial surgery. Results showed that there were significant differences between LUAD and LUSC in gene expression and microbial abundance associated with recurrence and metastasis. Compared with non-RM, the bacterial community of RM had a lower richness in LUSC. In LUSC, host genes significantly correlated with tissue microbe, whereas host-tissue microbe interaction in LUAD was rare. Then, we established a novel multimodal machine learning model based on genes and microbes to predict the recurrence and metastasis risk of a LUSC patient, which achieves an area under the curve (AUC) of 0.81. In addition, the predicted risk score was significantly associated with the patient's survival. IMPORTANCE Our study elucidates significant differences in RM-associated host-microbe interactions between LUAD and LUSC. Besides, the microbes in tumor tissue could be used to predict the RM risk of LUSC, and the predicted risk score is associated with patients' survival.

MDAlmc: A Novel Low-rank Matrix Completion Model for MiRNADisease Association Prediction by Integrating Similarities among MiRNAs and Diseases.

INTRODUCTION: The importance of microRNAs (miRNAs) has been emphasized by an increasing number of studies, and it is well-known that miRNA dysregulation is associated with a variety of complex diseases. Revealing the associations between miRNAs and diseases are essential to disease prevention, diagnosis, and treatment. METHODS: However, traditional experimental methods in validating the roles of miRNAs in diseases could be very expensive, labor-intensive and time-consuming. Thus, there is a growing interest in predicting miRNA-disease associations by computational methods. Though many computational methods are in this category, their prediction accuracy needs further improvement for downstream experimental validation. In this study, we proposed a novel model to predict miRNA-disease associations by low-rank matrix completion (MDAlmc) integrating miRNA functional similarity, disease semantic similarity, and known miRNA-disease associations. In the 5-fold cross-validation, MDAlmc achieved an average AUROC of 0.8709 and AUPRC of 0.4172, better than those of previous models. RESULTS: Among the case studies of three important human diseases, the top 50 predicted miRNAs of 96% (breast tumors), 98% (lung tumors), and 90% (ovarian tumors) have been confirmed by previous literatures. And the unconfirmed miRNAs were also validated to be potential disease-associated miRNAs. CONCLUSION: MDAlmc is a valuable computational resource for miRNA-disease association prediction.

MNNMDA: Predicting human microbe-disease association via a method to minimize matrix nuclear norm.

Identifying the potential associations between microbes and diseases is the first step for revealing the pathological mechanisms of microbe-associated diseases. However, traditional culture-based microbial experiments are expensive and time-consuming. Thus, it is critical to prioritize disease-associated microbes by computational methods for further experimental validation. In this study, we proposed a novel method called MNNMDA, to predict microbe-disease associations (MDAs) by applying a Matrix Nuclear Norm method into known microbe and disease data. Specifically, we first calculated Gaussian interaction profile kernel similarity and functional similarity for diseases and microbes. Then we constructed a heterogeneous information network by combining the integrated disease similarity network, the integrated microbe similarity network and the known microbe-disease bipartite network. Finally, we formulated the microbe-disease association prediction problem as a low-rank matrix completion problem, which was solved by minimizing the nuclear norm of a matrix with a few regularization terms. We tested the performances of MNNMDA in three datasets including HMDAD, Disbiome, and Combined Data with small, medium and large sizes respectively. We also compared MNNMDA with 5 state-of-the-art methods including KATZHMDA, LRLSHMDA, NTSHMDA, GATMDA, and KGNMDA, respectively. MNNMDA achieved area under the ROC curves (AUROC) of 0.9536 and 0.9364 respectively on HDMAD and Disbiome, better than the AUCs of compared methods under the 5-fold cross-validation for all microbe-disease associations. It also obtained a relatively good performance with AUROC 0.8858 in the combined data. In addition, MNNMDA was also better than other methods in area under precision and recall curve (AUPR) under the 5-fold cross-validation for all associations, and in both AUROC and AUPR under the 5-fold cross-validation for diseases and the 5-fold cross-validation for microbes. Finally, the case studies on colon cancer and inflammatory bowel disease (IBD) also validated the effectiveness of MNNMDA. In conclusion, MNNMDA is an effective method in predicting microbe-disease associations. Availability: The codes and data for this paper are freely available at Github https://github.com/Haiyan-Liu666/MNNMDA.

Single-cell technologies: From research to application.

In recent years, more and more single-cell technologies have been developed. A vast amount of single-cell omics data has been generated by large projects, such as the Human Cell Atlas, the Mouse Cell Atlas, the Mouse RNA Atlas, the Mouse ATAC Atlas, and the Plant Cell Atlas. Based on these single-cell big data, thousands of bioinformatics algorithms for quality control, clustering, cell-type annotation, developmental inference, cell-cell transition, cell-cell interaction, and spatial analysis are developed. With powerful experimental single-cell technology and state-of-the-art big data analysis methods based on artificial intelligence, the molecular landscape at the single-cell level can be revealed. With spatial transcriptomics and single-cell multi-omics, even the spatial dynamic multi-level regulatory mechanisms can be deciphered. Such single-cell technologies have many successful applications in oncology, assisted reproduction, embryonic development, and plant breeding. We not only review the experimental and bioinformatics methods for single-cell research, but also discuss their applications in various fields and forecast the future directions for single-cell technologies. We believe that spatial transcriptomics and single-cell multi-omics will become the next booming business for mechanism research and commercial industry.

Bacterial biomarkers capable of identifying recurrence or metastasis carry disease severity information for lung cancer.

Background: Local recurrence and distant metastasis are the main causes of death in patients with lung cancer. Multiple studies have described the recurrence or metastasis of lung cancer at the genetic level. However, association between the microbiome of lung cancer tissue and recurrence or metastasis remains to be discovered. Here, we aimed to identify the bacterial biomarkers capable of distinguishing patients with lung cancer from recurrence or metastasis, and how it related to the severity of patients with lung cancer. Methods: We applied microbiome pipeline to bacterial communities of 134 non-recurrence and non-metastasis (non-RM) and 174 recurrence or metastasis (RM) samples downloaded from The Cancer Genome Atlas (TCGA). Co-occurrence network was built to explore the bacterial interactions in lung cancer tissue of RM and non-RM. Finally, the Kaplan-Meier survival analysis was used to evaluate the association between bacterial biomarkers and patient survival. Results: Compared with non-RM, the bacterial community of RM had lower richness and higher Bray-Curtis dissimilarity index. Interestingly, the co-occurrence network of non-RM was more complex than RM. The top 500 genera in relative abundance obtained an area under the curve (AUC) of 0.72 when discriminating between RM and non-RM. There were significant differences in the relative abundances of Acidovorax, Clostridioides, Succinimonas, and Shewanella, and so on between RM and non-RM. These biomarkers played a role in predicting the survival of lung cancer patients and were significantly associated with lung cancer stage. Conclusion: This study provides the first evidence for the prediction of lung cancer recurrence or metastasis by bacteria in lung cancer tissue. Our results highlights that bacterial biomarkers that distinguish RM and non-RM are also associated with patient survival and disease severity.

ICSDA: a multi-modal deep learning model to predict breast cancer recurrence and metastasis risk by integrating pathological, clinical and gene expression data.

Breast cancer patients often have recurrence and metastasis after surgery. Predicting the risk of recurrence and metastasis for a breast cancer patient is essential for the development of precision treatment. In this study, we proposed a novel multi-modal deep learning prediction model by integrating hematoxylin & eosin (H&E)-stained histopathological images, clinical information and gene expression data. Specifically, we segmented tumor regions in H&E into image blocks (256 × 256 pixels) and encoded each image block into a 1D feature vector using a deep neural network. Then, the attention module scored each area of the H&E-stained images and combined image features with clinical and gene expression data to predict the risk of recurrence and metastasis for each patient. To test the model, we downloaded all 196 breast cancer samples from the Cancer Genome Atlas with clinical, gene expression and H&E information simultaneously available. The samples were then divided into the training and testing sets with a ratio of 7: 3, in which the distributions of the samples were kept between the two datasets by hierarchical sampling. The multi-modal model achieved an area-under-the-curve value of 0.75 on the testing set better than those based solely on H&E image, sequencing data and clinical data, respectively. This study might have clinical significance in identifying high-risk breast cancer patients, who may benefit from postoperative adjuvant treatment.

Predicting colorectal cancer tumor mutational burden from histopathological images and clinical information using multi-modal deep learning.

MOTIVATION: Tumor mutational burden (TMB) is an indicator of the efficacy and prognosis of immune checkpoint therapy in colorectal cancer (CRC). In general, patients with higher TMB values are more likely to benefit from immunotherapy. Though whole-exome sequencing is considered the gold standard for determining TMB, it is difficult to be applied in clinical practice due to its high cost. There are also a few DNA panel-based methods to estimate TMB; however, their detection cost is also high, and the associated wet-lab experiments usually take days, which emphasize the need for faster and cheaper alternatives. RESULTS: In this study, we propose a multi-modal deep learning model based on a residual network (ResNet) and multi-modal compact bilinear pooling to predict TMB status (i.e. TMB high (TMB_H) or TMB low(TMB_L)) directly from histopathological images and clinical data. We applied the model to CRC data from The Cancer Genome Atlas and compared it with four other popular methods, namely, ResNet18, ResNet50, VGG19 and AlexNet. We tested different TMB thresholds, namely, percentiles of 10%, 14.3%, 15%, 16.3%, 20%, 30% and 50%, to differentiate TMB_H and TMB_L.For the percentile of 14.3% (i.e. TMB value 20) and ResNet18, our model achieved an area under the receiver operating characteristic curve of 0.817 after 5-fold cross-validation, which was better than that of other compared models. In addition, we also found that TMB values were significantly associated with the tumor stage and N and M stages. Our study shows that deep learning models can predict TMB status from histopathological images and clinical information only, which is worth clinical application.

Evaluating the Microsatellite Instability of Colorectal Cancer Based on Multimodal Deep Learning Integrating Histopathological and Molecular Data.

Microsatellite instability (MSI), an important biomarker for immunotherapy and the diagnosis of Lynch syndrome, refers to the change of microsatellite (MS) sequence length caused by insertion or deletion during DNA replication. However, traditional wet-lab experiment-based MSI detection is time-consuming and relies on experimental conditions. In addition, a comprehensive study on the associations between MSI status and various molecules like mRNA and miRNA has not been performed. In this study, we first studied the association between MSI status and several molecules including mRNA, miRNA, lncRNA, DNA methylation, and copy number variation (CNV) using colorectal cancer data from The Cancer Genome Atlas (TCGA). Then, we developed a novel deep learning framework to predict MSI status based solely on hematoxylin and eosin (H&E) staining images, and combined the H&E image with the above-mentioned molecules by multimodal compact bilinear pooling. Our results showed that there were significant differences in mRNA, miRNA, and lncRNA between the high microsatellite instability (MSI-H) patient group and the low microsatellite instability or microsatellite stability (MSI-L/MSS) patient group. By using the H&E image alone, one can predict MSI status with an acceptable prediction area under the curve (AUC) of 0.809 in 5-fold cross-validation. The fusion models integrating H&E image with a single type of molecule have higher prediction accuracies than that using H&E image alone, with the highest AUC of 0.952 achieved when combining H&E image with DNA methylation data. However, prediction accuracy will decrease when combining H&E image with all types of molecular data. In conclusion, combining H&E image with deep learning can predict the MSI status of colorectal cancer, the accuracy of which can further be improved by integrating appropriate molecular data. This study may have clinical significance in practice.