15Head and neck squamous cell carcinoma (HNSCC) are a heterogeneous group of cancers 16 affecting multiple subsites, including oral cavity. Oral or anterior tongue tumors (OTSCC) are an 17 aggressive group of squamous cell carcinomas, characterized by their early spread to lymph nodes 18 and higher rate of regional failure compared to other oral cavity cancers. There is a rise in the 19 incidence of oral tongue cancer among younger population (<50yrs); many of who lack the typical 20 associated risk factors of alcohol and/or tobacco exposure. In order to carry out an ensemble 21 learning and prediction method with multiple parameters classifying survival, we generated data, on 22 somatic mutations in genes from exome sequencing, immediate upstream and downstream flanking 23 nucleotides of the somatic mutations, DNA methylation, loss of heterozygosity (LOH), copy 24 number variations (CNV), gene expression, significant pathways altered and Human Papilloma 25 Virus (HPV) infection, from 50 OTSCC patients. Results of our analysis identified somatic 26 mutations in NOTCH2 and/or TP53, and/or LOH in 11p to associate with better disease free 27 survival in HPV positive patients (P = 0.0254) and not in HPV negative patients (P = 0.414). We 28 validated the latter in patients without HPV infection from TCGA cohort (P = 0.369, N = 17 for 29 2 TCGA_ OralTongue; P = 0.472, N = 67 for all TCGA_HNSCC patients). Integrated analysis, 30 including pathways, linked survival with apoptosis and aberrant methylation in SLC38A8 (P = 31 0.0129). 32 33 Author Summary 34 Oral tongue squamous cell carcinomas (OTSCC) are a homogenous group of head and neck 35 tumors characterized with aggressive behavior among younger patients. In this report, we have 36 analysed genetic variants, expression and DNA methylation changes across 50 oral tongue primary 37 tumors along with the Human Papilloma Virus (HPV) infection status in those tumors to identify 38 factors associated with disease free survival. Our data identified somatic mutations in the genes 39 NOTCH2, TP53 and LOH in 11p, to be significantly associated with better disease free survival in 40 HPV positive patients (P = 0.0254), but not in HPV negative patients (P = 0.414). We validated the 41 latter using patients without HPV infection from TCGA (P = 0.369, N = 17 for 42 TCGA_OralTongue; P = 0.472, N = 67 for all TCGA_HNSCC patients). Integrated analysis linked 43 survival with apoptosis and aberrant methylation in SLC38A8 (P = 0.0129).44 45 48 cause of cancer worldwide [1]. In India, they account for almost 30% of all cancer cases [2]. Studies 49 on molecular biology of HNSCC in the past 5yrs have largely been concentrated on cataloging 50various genetic changes in many cancer types using high-throughput sequencing assays and 51 computational methods [3][4][5][6][7]. Unlike other oral cavity subsites, squamous cell carcinomas of 52 oral/anterior tongue tend to be associated more with younger patients [8, 9], early spread to lymph 53 nodes [8] and a higher regional failure compared t...

Plant-derived secondary metabolites play a vital role in the food, pharmaceutical, agrochemical and cosmetic industry. Metabolite concentrations are measured after extraction, biochemistry and analyses, requiring time, access to expensive equipment, reagents and specialized skills. Additionally, metabolite concentration often varies widely among plants, even within a small area. A quick method to estimate the metabolite concentration class (high or low) will significantly help in selecting trees yielding high metabolites for the metabolite production process. Here, we demonstrate a deep learning approach to estimate the concentration class of an intracellular metabolite, azadirachtin, using models built with images of leaves and fruits collected from randomly selected Azadirachta indica (neem) trees in an area spanning >500,000 sqkms and their corresponding biochemically measured metabolite concentrations. We divided the input data randomly into training-and test-sets ten times to avoid sampling bias and to optimize the model parameters during cross-validation. The training-set contained >83,000 fruit and >86,000 leaf images. The best models yielded prediction errors of 19.13% and 15.11% (for fruit), and 8% and 26.67% (for leaf), each, for low and high metabolite classes, respectively. We further validated the fruit model using independently collected fruit images from different locations spanning nearly 130,000 sqkms, with 70% accuracy. We developed a desktop application to scan offline image(s) and a mobile application for real-time utility to predict the metabolite content class. Our work demonstrates the use of a deep learning method to estimate the concentration class of an intracellular metabolite using images, and has broad applications and utility.Key words: convolutional neural networks (CNN); image-based deep learning; intracellular metabolite content class; artificial intelligence; azadirachtin BackgroundMeasuring the concentration of an analyte, such as metabolite, enzyme, protein or any other chemical moiety within plants, animals and microbial cells is a frequent practice in biology. To do this, chemical, biochemical, immunological or imaging-based methods are usually followed, where each method provides a different type of readout.Although accurate and precise, the currently employed analytical methods require extensive sample handling and preparation time, expensive reagents and equipment, and specialized skills. In circumstances where knowing the exact concentration of the intracellular metabolite is not necessary, and a quick and rough estimate of its concentration class (either high or low) is enough, use of conventional measurement methods although unnecessary is currently unavoidable. A method that provides a quick readout of the analyte concentration class, however ideal in such cases, is currently not available. Metabolites, primary and secondary, are intermediate products of metabolic reactions catalyzed by enzymes. Examples of some primary metabolites are amino acids, vitamins, organic ac...