Biology subjects

Krishnan, N. M.

Publications and source records attributed to Krishnan, N. M..

12 recordsLinked to original sources

bioRxiv2026

X-Plat: A polynomial regression based tool for cross-platform transformation of expression and methylation data

Many biological studies could benefit from combining data from legacy microarray and high-throughput sequencing platforms, especially in clinical domains where collecting additional samples is not possible. However, incompatibility between platforms makes legacy data difficult to integrate, owing to differences in platform design, target preparation, and dependence on prior annotations. Here, we describe X-Plat, a cross-platform data transformation tool for both expression and methylation assays that inter-converts data between microarray and sequencing platforms using per-gene second-degree polynomial regression. X-Plat learns cross-platform conversion rules from paired microarray-sequencing datasets spanning multiple conditions, sample sources, organisms, and platforms, and evaluates performance using cross-validated root mean square error (RMSE) per gene. In rat, Arabidopsis, and human datasets, X-Plat achieved lower cross-validated RMSE than TDM, HARMONY, and HARMONY2 for the vast majority of genes ([≥]95% in all sequencing-to-array transformations and most array-to-sequencing transformations, with [~]82% in the Arabidopsis array-to-sequencing setting), and these findings were confirmed using RMSE on held-out test samples from the first cross-validation fold. X-Plat also achieved low RMSE ([≤]0.2) for the majority of CpG regions in paired human array-sequencing methylation datasets. Using X-Plat, users can transform data between microarray and high-throughput sequencing platforms, enabling cross-platform comparison and reuse of legacy cohorts. GRAPHICAL ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=84 SRC="FIGDIR/small/707273v2_ufig1.gif" ALT="Figure 1"> View larger version (19K): org.highwire.dtl.DTLVardef@30cabeorg.highwire.dtl.DTLVardef@14ea2eborg.highwire.dtl.DTLVardef@1aa232forg.highwire.dtl.DTLVardef@ebf4f3_HPS_FORMAT_FIGEXP M_FIG C_FIG

genomics↗

bioRxiv2024

Fruit-In-Sight: a deep learning-based framework for secondary metabolite class prediction using fruit and leaf images

Fruits produce a wide variety of secondary metabolites of great economic value. Analytical measurement of secondary metabolites is tedious, time-consuming and expensive. Additionally, metabolite concentration varies greatly from tree to tree, making it difficult to choose trees for fruit collection. The current study tested whether deep learning-based models can be developed using fruit and leaf images alone to predict a metabolites concentration class (high or low). We collected fruits and leaves (n = 1045) from neem trees grown in the wild across 0.6 million sq km, imaged those, measured concentration of five metabolites (azadirachtin, deacetyl-salannin, salannin, nimbin and nimbolide) using high-performance liquid chromatography and used those to train deep learning models for metabolite class prediction. The best model out of the seven tested (YOLOv5, GoogLeNet, InceptionNet, EfficientNet_B0, Resnext_50, Resnet18, and SqueezeNet) provided a validation F1 score of 0.93 and a test F1 score of 0.88. The sensitivity and specificity of the fruit model alone in the test set were 83.52 {+/-} 6.19 and 82.35 {+/-} 5.96 and 79.40 {+/-} 8.50 and 85.64 {+/-} 6.21, for the low and the high class, respectively. The sensitivity was further boosted to 92.67{+/-} 5.25 for the low class and 88.11 {+/-} 9.17 for the high class and the specificity to 100% for both classes, using a multi-analyte framework. We incorporated the model in an Android mobile App Fruit-In-Sight that uses fruit and leaf images to decide whether to pick or not pick the fruits from a specific tree based on the metabolite concentration class. Our study provides evidence that images of fruits and leaves alone can predict the concentration class of a secondary metabolite without using extensive analytical laboratory procedures and equipment and makes the process of choosing the right tree for fruit collection easy and free of equipment and additional cost.

bioinformatics↗

bioRxiv2023

Building flexible and robust analysis frameworks for molecular subtyping of cancers

Molecular subtyping is essential to infer tumor aggressiveness and predict prognosis. In practice, tumor profiling requires in-depth knowledge of bioinformatics tools involved in the processing and analysis of the generated data. Additionally, data incompatibility (e.g., microarray vs. RNA sequencing data) and technical and uncharacterized biological variance between training and test data can pose challenges in classifying individual samples. In this article, we provide a roadmap for implementing bioinformatics frameworks for molecular profiling of human cancers in a clinical diagnostic setting. We describe a framework for integrating several methods for quality control, normalization, batch correction, classification, and reporting and a use case of the framework in breast cancer.

bioinformatics↗

bioRxiv2022

Single and multi-analyte deep learning-based analysis framework for class prediction in biological images

Measurement of biological analytes, characterizing flavor in fruits, is a cumbersome, expensive and time-consuming process. Fruits with higher concentration of analytes have greater commercial or nutritional values. Here, we tested a deep learning-based framework with fruit images to predict the class (sweet or sour and high or low) of analytes using images from two types of trees in a single and multi-analyte mode. We used fruit images from kinnow (n = 3,451), an edible hybrid mandarin and neem (n = 1,045), a tree with agrochemical and pharmaceutical properties. We measured sweetness in kinnows and five secondary metabolites in neem fruits (azadirachtin or A, deacetyl-salannin or D, salannin or S, nimbin or N and nimbolide or E) using a refractometer and high-performance liquid chromatography, respectively. We trained the models for 300 epochs, before and after hyper-parameters evolution, using 300 generations with 50 epochs/generation, estimated the best models and evaluated their performance on 10% of independent images. The validation F1score and test accuracies were 0.79 and 0.77, and 82.55% and 60.8%, respectively for kinnow and neem A analyte. A multi-analyte model enhanced the neem A models prediction to high class when the D:N:Ss combined class predictions were high:low:high and to low class when D:Ns combined class predictions were low:high respectively. The test accuracy increased further to ~70% with a 10-fold cross-validation error of 0.257 across ten randomly split train:validation:test sets proving the potential of a multi-analyte model to enhance the prediction accuracy, especially when the numbers of images are limiting.

plant biology↗

bioRxiv2022

Photosystem genes in chloroplast and their interacting partners: A case for molecular adaptation to carnivory in Nepenthaceae.

Molecular adaptations are prevalent in carnivorous plants in response to habitat and environmental stress. We used the chloroplast genome and characterized the specific adaptations in the photosystem genes and their interacting partners in Nepenthes khasiana, a carnivorous pitcher plant. When compared with the carnivorous and non-carnivorous groups across Caryophyllales, Lamiales, Poales, Ericales, and Oxalidales, we found Nepenthes-specific changes in psaA, psaB, psaC and psaH. Of these, only a single amino acid change each, G147 in the protein psaA and R40 in the protein ndhD, impacted the three-dimensional structural conformation of the corresponding proteins. Modeling the interaction between the psaA and the ndhD proteins identified group-specific changes between the models between Nepentheceae versus others. The least distance between the structure-impacting residues of psaA and ndhD was 25.9 [A] for Nepenthes and 19.4 [A] for non-Nepenthes models. Given that the chloroplast ndh and photosystem I subunits form a large super-complex with the light-harvesting carrier proteins from the nucleus to mediate cyclic electron transport, our observations may indicate specific adaptations in the cyclic electron transport arm of the photosynthetic machinery in the Nepenthes species.

bioinformatics↗

bioRxiv2018

Prediction of a plant intracellular metabolite content class using image-based deep learning

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.

bioinformatics↗

bioRxiv2018

Functional Genomics Screen with Pooled shRNA Library and Gene Expression Profiling with Extracts of Azadirachta indica Identify Potential Pathways for Therapeutic Targets in Head and Neck Squamous Cell Carcinoma

Tumor suppression by the extracts of Azadirachta indica (neem) works via anti-proliferation, cell cycle arrest, and apoptosis, demonstrated previously using cancer cell lines and live animal models. However, very little is known about the molecular targets and pathways that the neem extracts and the associated compounds act through. Here, we address this using a genome-wide functional pooled shRNA screen on head and neck squamous cell carcinoma cell line treated with crude neem leaf extracts, known for their anti-tumorigenic activity. By analyzing differences in global clonal sizes of the shRNA-infected cells cultured under no treatment and treatment with neem leaf extract conditions, assayed using next-generation sequencing, we found 225 genes affected the cancer cell growth in the shRNA-infected cells treated with neem extract. Pathway enrichment analyses of whole-genome gene expression data from cells temporally treated with neem extract revealed important roles played by the TGF-{beta} pathway and HSF-1-related gene network. Our results indicate that neem extract simultaneously affects various important molecular signaling pathways in head and neck cancer cells, some of which may be therapeutic targets for this devastating tumor.

genomics↗

bioRxiv2017

Comparative analyses of venom-associated genes from an Old World viper, Daboia russelii

Molecular basis of toxin gene diversity among snakes is poorly understood. Lack of whole genome sequence information for most snakes makes studies on toxin genes and their orthologous counterparts difficult. One of the challenges in studying snake genomes is the acquisition of biological material from live animals, especially from the venomous ones. Additionally, in certain geographies, Government permission is required to handle live snakes making the process cumbersome and time-consuming. Here, we report comparative sequence analyses of toxin genes from Russells viper (Daboia russelii) using whole-genome sequencing data obtained from the skin exuviate. In addition to the comparative analyses of 46 toxin-associated proteins, we present evidence of unique sequence motifs in five key toxin-associated protein domains; nerve growth factor (NGF), platelet derived growth factor (PDGF), Kunitz/Bovine pancreatic trypsin inhibitor (Kunitz BPTI), cysteine-rich secretory proteins, antigen 5, and pathogenesis-related 1 proteins (CAP) and cysteine-rich secretory protein (CRISP). We compared the venom-associated domains from Russells viper with those from both venomous and non-venomous vertebrates and invertebrates. The in silico study on structures identified V11 and T35 in the NGF domain; F23 and A29 in the PDGF domain; N69, K2 and A5 in the CAP domain; and Q17 in the CRISP domain to be responsible for differences in the largest pockets across the protein domain structures in New World vipers, Old World vipers and elapids. Similarly, residues F10, Y11 and E20 appear to play an important role in the protein structures across the kunitz protein domain of viperids and elapids. Our study sheds light on the uniqueness of these key toxin-associated proteins and their evolution in vipers.\n\nData deposition: Russells viper sequence data is deposited in the NCBI SRA database under the accession number SRR5506741 and the GenBank accession numbers for the individual venom-associated genes is provided in Table S1.

bioinformatics↗

bioRxiv2017

A minimal set of internal control genes for gene expression studies in head and neck squamous cell carcinoma

BackgroundSelection of the right reference gene(s) is crucial in the analysis and interpretation of gene expression data. In head and neck cancer, studies evaluating the efficacy of internal reference genes are rare. Here, we present data for a minimal set of candidates as internal control genes for gene expression studies in head and neck cancer.\n\nMethodsWe analyzed data from multiple sources (in house whole-genome gene expression microarrays, n=21; TCGA RNA-seq, n=42, and published gene expression studies in head and neck tumors from literature) to come up with a set of genes (discovery set) for their stable expression across tumor and normal tissues. We then performed independent validation of their expression using qPCR in 14 tumor:normal pairs. Genes in the discovery set were ranked using four different algorithms (BestKeeper, geNorm, NormFinder, and comparative delta Ct) and a web-based comparative tool, RefFinder, for their stability and variance in expression across tissues.\n\nResultsOur analyses resulted in 18 genes (discovery set) that had lowest variance and high level of expression across tumor and normal samples. Independent experimental validation and analyses with multiple tools resulted in top ranked five genes (RPL30, RPL27, PSMC5, OAZ1 and MTCH1) out of which, RPL30 (60S ribosomal protein L30) and RPL27 (60S ribosomal protein L27), performed best and were abundantly expressed across tumor and normal tissues.\n\nConclusionsRPL30 and RPL27 are stably expressed in HNSCC and should be used as internal control genes in gene expression in head and neck tumors studies.

cancer biology↗

bioRxiv2017

CAFE MOCHA: An Integrated Platform for Discovering Clinically Relevant Molecular Changes in Cancer; an Example of Distant Metastasis and Recurrence-linked Classifiers in Head and Neck Squamous Cell Carcinoma

BackgroundCAFE MOCHA (Clinical Association of Functionally Established MOlecular CHAnges) is an integrated GUI-driven computational and statistical framework to discover molecular signatures linked to a specific clinical attribute in a cancer type. We tested CAFE MOCHA in head and neck squamous cell carcinoma (HNSCC) for discovering a signature linked to distant metastasis and recurrence (MR) in 517 tumors from TCGA and validated the signature in 18 tumors from an independent cohort.\n\nMethodsThe platform integrates mutations and indels, gene expression, DNA methylation and copy number variations to discover a classifier first, predict an incoming tumour for the same by pulling defined class variables into a single framework that incorporates a coordinate geometry-based algorithm, called Complete Specificity Margin Based Clustering (CSMBC) with 100% specificity. CAFE MOCHA classifies an incoming tumour sample using either a matched normal or a built-in database of normal tissues. The application is packed and deployed using the install4j multi-platform installer.\n\nResultsWe tested CAFE MOCHA to discover a signature for distant metastasis and recurrence in HNSCC. The signature MR44 in HNSCC yielded 80% sensitivity and 100% specificity in the discovery stage and 100% sensitivity and 100% specificity in the validation stage.\n\nConclusionsCAFE MOCHA is a cancer type- and clinical attribute-agnostic computational and statistical framework to discover integrated molecular signature for a specific clinical attribute.\n\nCAFE MOCHA is available in GitHub (https://github.com/binaypanda/CAFEMOCHA).

cancer biology↗

bioRxiv2016

RNAtor: an Android-based application for biologists to plan RNA sequencing experiments.

RNA sequencing (RNA-seq) is a powerful technology for identification of novel transcripts (coding, non-coding and splice variants), understanding of transcript structures and estimation of gene and/or allelic expression. There are specific challenges that biologists face in determining the number of replicates to use, total number of sequencing reads to generate for detecting marginally differentially expressed transcripts and the number of lanes in a sequencing flow cell to use for the production of right amount of information. Although past studies attempted answering some of these questions, there is a lack of accessible and biologist-friendly mobile applications to answer these questions. Keeping this in mind, we have developed RNAtor, a mobile application for Android platforms, to aid biologists in correctly designing their RNA-seq experiments. The recommendations from RNAtor are based on simulations and real data.\n\nAvailability and ImplementationThe Android version of RNAtor is available on Google Play Store and the code from GitHub (https://github.com/binaypanda/RNAtor).

bioinformatics↗

bioRxiv2016

High-risk human papillomavirus in oral cavity squamous cell carcinoma

PurposeThe prevalence of human papillomavirus (HPV) in oral cavity squamous cell carcinoma (OSCC) varies significantly based on assay sensitivity and patient geography. Accurate detection is essential to understand the role of HPV in disease prognosis and management of patients with OSCC.\n\nMethodsWe generated and integrated data from multiple analytes (HPV DNA, HPV RNA, and p16), assays (immunohistochemistry, PCR, qPCR and digital PCR) and molecular changes (somatic mutations and DNA methylation) from 153 OSCC patients to correlate p16 expression, HPV DNA, and HPV RNA with HPV incidence and patient survival.\n\nResultsHigh prevalence (33-58%) of HPV16/18 DNA did not correlate with the presence of transcriptionally active viral genomes (15%) in tumors. Eighteen percent of the tumors were p16 positive. and only 6% were both HPV DNA and RNA positive. Most tumors with relatively high-copy HPV DNA, and/or HPV RNA, but not with HPV DNA alone (irrespective of copy number), were wild-type for TP53 and CASP8 genes. In our study, p16 protein, HPV DNA and HPV RNA, either alone or in combinations, did not correlate with patient survival. Nine HPV-associated genes stratified the virus +ve from the -ve tumor group with high confidence (p<0.008) when HPV DNA copy number and/or HPV RNA were considered to define HPV positivity and not HPV DNA alone irrespective of their copy number (p < 0.2).\n\nConclusionsIn OSCC, the presence of both HPV RNA and p16 are rare. HPV DNA alone is not an accurate measure of HPV positivity and therefore not informative. Moreover, HPV DNA, RNA or p16 dont correlate with outcome.

cancer biology↗