Predicting herb-disease associations using network-based measures in human protein interactome.

BACKGROUND: Natural herbs are frequently used to treat diseases or to relieve symptoms in many countries. Moreover, as their safety has been proven for a long time, they are considered as main sources of new drug development. However, in many cases, the herbs are still prescribed relying on ancient records and/or traditional practices without scientific evidences. More importantly, the medicinal efficacy of the herbs has to be evaluated in the perspective of MCMT (multi-compound multi-target) effects, but most efforts focus on identifying and analyzing a single compound experimentally. To overcome these hurdles, computational approaches which are based on the scientific evidences and are able to handle the MCMT effects are needed to predict the herb-disease associations. RESULTS: In this study, we proposed a network-based in silico method to predict the herb-disease associations. To this end, we devised a new network-based measure, WACP (weighted average closest path length), which not only quantifies proximity between herb-related genes and disease-related genes but also considers compound compositions of each herb. As a result, we confirmed that our method successfully predicts the herb-disease associations in the human protein interactome (AUROC = 0.777). In addition, we observed that our method is superior than the other simple network-based proximity measures (e.g. average shortest and closest path length). Additionally, we analyzed the associations between Brassica oleracea var. italica and its known associated diseases more specifically as case studies. Finally, based on the prediction results of the WACP, we suggested novel herb-disease pairs which are expected to have potential relations and their literature evidences. CONCLUSIONS: This method could be a promising solution to modernize the use of the natural herbs by providing the scientific evidences about the molecular associations between the herb-related genes targeted by multiple compounds and the disease-related genes in the human protein interactome.

CancerGATE: Prediction of cancer-driver genes using graph attention autoencoders.

Discovery of the cancer type specific-driver genes is important for understanding the molecular mechanisms of each cancer type and for providing proper treatment. Recently, graph deep learning methods became widely used in finding cancer-driver genes. However, previous methods had limited performance in individual cancer types due to a small number of cancer-driver genes used in training and biases toward the cancer-driver genes used in training the models. Here, we introduce a novel pipeline, CancerGATE that predicts the cancer-driver genes using graph attention autoencoder (GATE) to learn in a self-supervised manner and can be applied to each of the cancer types. CancerGATE utilizes biological network topology and multi-omics data from 15 types of cancer of 20,079 samples from the cancer genome atlas (TCGA). Attention coefficients calculated in the model are used to prioritize cancer-driver genes by comparing coefficients of cancer and normal contexts. CancerGATE shows a higher AUPRC with a difference ranging from 1.5 % to 36.5 % compared to the previous graph deep learning models in each cancer type. We also show that CancerGATE is free from the bias toward cancer-driver genes used in training, revealing mechanisms of the cancer-driver genes in specific cancer types. Finally, we propose novel cancer-driver gene candidates that could be therapeutic targets for specific cancer types.

Community cohesion looseness in gene networks reveals individualized drug targets and resistance.

Community cohesion plays a critical role in the determination of an individual's health in social science. Intriguingly, a community structure of gene networks indicates that the concept of community cohesion could be applied between the genes as well to overcome the limitations of single gene-based biomarkers for precision oncology. Here, we develop community cohesion scores which precisely quantify the community ability to retain the interactions between the genes and their cellular functions in each individualized gene network. Using breast cancer as a proof-of-concept study, we measure the community cohesion score profiles of 950 case samples and predict the individualized therapeutic targets in 2-fold. First, we prioritize them by finding druggable genes present in the community with the most and relatively decreased scores in each individual. Then, we pinpoint more individualized therapeutic targets by discovering the genes which greatly contribute to the community cohesion looseness in each individualized gene network. Compared with the previous approaches, the community cohesion scores show at least four times higher performance in predicting effective individualized chemotherapy targets based on drug sensitivity data. Furthermore, the community cohesion scores successfully discover the known breast cancer subtypes and we suggest new targeted therapy targets for triple negative breast cancer (e.g. KIT and GABRP). Lastly, we demonstrate that the community cohesion scores can predict tamoxifen responses in ER+ breast cancer and suggest potential combination therapies (e.g. NAMPT and RXRA inhibitors) to reduce endocrine therapy resistance based on individualized characteristics. Our method opens new perspectives for the biomarker development in precision oncology.

Identifying prognostic subgroups of luminal-A breast cancer using deep autoencoders and gene expressions.

Luminal-A breast cancer is the most frequently occurring subtype which is characterized by high expression levels of hormone receptors. However, some luminal-A breast cancer patients suffer from intrinsic and/or acquired resistance to endocrine therapies which are considered as first-line treatments for luminal-A breast cancer. This heterogeneity within luminal-A breast cancer has required a more precise stratification method. Hence, our study aims to identify prognostic subgroups of luminal-A breast cancer. In this study, we discovered two prognostic subgroups of luminal-A breast cancer (BPS-LumA and WPS-LumA) using deep autoencoders and gene expressions. The deep autoencoders were trained using gene expression profiles of 679 luminal-A breast cancer samples in the METABRIC dataset. Then, latent features of each samples generated from the deep autoencoders were used for K-Means clustering to divide the samples into two subgroups, and Kaplan-Meier survival analysis was performed to compare prognosis (recurrence-free survival) between them. As a result, the prognosis between the two subgroups were significantly different (p-value = 5.82E-05; log-rank test). This prognostic difference between two subgroups was validated using gene expression profiles of 415 luminal-A breast cancer samples in the TCGA BRCA dataset (p-value = 0.004; log-rank test). Notably, the latent features were superior to the gene expression profiles and traditional dimensionality reduction method in terms of discovering the prognostic subgroups. Lastly, we discovered that ribosome-related biological functions could be potentially associated with the prognostic difference between them using differentially expressed genes and co-expression network analysis. Our stratification method can be contributed to understanding a complexity of luminal-A breast cancer and providing a personalized medicine.

Mutation of SPOTTED LEAF3 (SPL3) impairs abscisic acid-responsive signalling and delays leaf senescence in rice.

Lesion mimic mutants commonly display spontaneous cell death in pre-senescent green leaves under normal conditions, without pathogen attack. Despite molecular and phenotypic characterization of several lesion mimic mutants, the mechanisms of the spontaneous formation of cell death lesions remain largely unknown. Here, the rice lesion mimic mutant spotted leaf3 (spl3) was examined. When grown under a light/dark cycle, the spl3 mutant appeared similar to wild-type at early developmental stages, but lesions gradually appeared in the mature leaves close to heading stage. By contrast, in spl3 mutants grown under continuous light, severe cell death lesions formed in developing leaves, even at the seedling stage. Histochemical analysis showed that hydrogen peroxide accumulated in the mutant, likely causing the cell death phenotype. By map-based cloning and complementation, it was shown that a 1-bp deletion in the first exon of Oryza sativa Mitogen-Activated Protein Kinase Kinase Kinase1 (OsMAPKKK1)/OsEDR1/OsACDR1 causes the spl3 mutant phenotype. The spl3 mutant was found to be insensitive to abscisic acid (ABA), showing normal root growth in ABA-containing media and delayed leaf yellowing during dark-induced and natural senescence. Expression of ABA signalling-associated genes was also less responsive to ABA treatment in the mutant. Furthermore, the spl3 mutant had lower transcript levels and activities of catalases, which scavenge hydrogen peroxide, probably due to impairment of ABA-responsive signalling. Finally, a possible molecular mechanism of lesion formation in the mature leaves of spl3 mutant is discussed.

Arabidopsis STAY-GREEN2 is a negative regulator of chlorophyll degradation during leaf senescence.

Chlorophyll (Chl) degradation causes leaf yellowing during senescence or under stress conditions. For Chl breakdown, STAY-GREEN1 (SGR1) interacts with Chl catabolic enzymes (CCEs) and light-harvesting complex II (LHCII) at the thylakoid membrane, possibly to allow metabolic channeling of potentially phototoxic Chl breakdown intermediates. Among these Chl catabolic components, SGR1 acts as a key regulator of leaf yellowing. In addition to SGR1 (At4g22920), the Arabidopsis thaliana genome contains an additional homolog, SGR2 (At4g11910), whose biological function remains elusive. Under senescence-inducing conditions, SGR2 expression is highly up-regulated, similarly to SGR1 expression. Here we show that SGR2 function counteracts SGR1 activity in leaf Chl degradation; SGR2-overexpressing plants stayed green and the sgr2-1 knockout mutant exhibited early leaf yellowing under age-, dark-, and stress-induced senescence conditions. Like SGR1, SGR2 interacted with LHCII but, in contrast to SGR1, SGR2 interactions with CCEs were very limited. Furthermore, SGR1 and SGR2 formed homo- or heterodimers, strongly suggesting a role for SGR2 in negatively regulating Chl degradation by possibly interfering with the proposed CCE-recruiting function of SGR1. Our data indicate an antagonistic evolution of the functions of SGR1 and SGR2 in Arabidopsis to balance Chl catabolism in chloroplasts with the dismantling and remobilizing of other cellular components in senescing leaf cells.