A5025134890- Machine Learning in Bioinformatics
- Protein Structure and Dynamics
- Computational Drug Discovery Methods
- Bioinformatics and Genomic Networks
- Visual Attention and Saliency Detection
- Face Recognition and Perception
- Neural dynamics and brain function
- Neuroscience of respiration and sleep
- Advanced Neural Network Applications
- Neonatal and fetal brain pathology
- Plant nutrient uptake and metabolism
- Neonatal Respiratory Health Research
- Cell Image Analysis Techniques
- Image Processing Techniques and Applications
- Domain Adaptation and Few-Shot Learning
- Microbial Metabolites in Food Biotechnology
- Genetic Mapping and Diversity in Plants and Animals
- Oral and gingival health research
- Biomedical Text Mining and Ontologies
- Enzyme Structure and Function
- Receptor Mechanisms and Signaling
- Head and Neck Cancer Studies
- Plant Molecular Biology Research
- Genomics and Phylogenetic Studies
- Speech and dialogue systems
Northeastern University
2020-2023
John Brown University
2021
Brown University
2021
Nationwide Children's Hospital
2017-2020
Universidad del Noreste
2020
Battelle
2017-2019
The Ohio State University
2018-2019
Dr. Reddy's Laboratories (Spain)
2018
Cerebral organoids (COs) are rapidly accelerating the rate of translational neuroscience based on their potential to model complex features developing human brain. Several studies have examined electrophysiological and neural network COs; however, no study has comprehensively investigated developmental trajectory properties in whole-brain COs correlated these with developmentally linked morphological cellular features. Here, we profiled neuroelectrical activities over span 5 months a...
Hypothetical proteins [HP] are those that predicted to be expressed in an organism, but no evidence of their existence is known. In the recent past, annotation and curation efforts have helped overcome challenge understanding diverse functions. Techniques decipher sequence-structure-function relationship, especially terms functional modelling HPs been developed by researchers, using features as classifiers for has not attempted. With rise number strategies, next-generation sequencing methods...
The improved competence of generative models can help building multi-modal virtual assistants that leverage modalities beyond language. By observing humans performing multi-step tasks, one build have situational awareness actions and tasks being performed, enabling them to cater assistance based on this understanding. In paper, we develop a Context-aware Instructional Task Assistant with Multi-modal Large Language Models (InsTALL) leverages an online visual stream (e.g. user's screen share...
Cereals are key contributors to global food security. Genes involved in the uptake (transport), assimilation and utilization of macro- micronutrients responsible for presence these nutrients grain straw. Although many genomic databases cereals available, there is currently no cohesive web resource manually curated nutrient use efficiency (NtUE)-related genes quantitative trait loci (QTLs). In this study, we present a web-resource containing information on NtUE-related genes/QTLs...
Delayed cancer detection is one of the common causes poor prognosis in case many cancers, including cancers oral cavity. Despite improvement and development new efficient gene therapy treatments, very little has been carried out to algorithmically assess impedance these carcinomas. In this work, from attributes or NCBI's datasets, viz. (i) name, (ii) gene(s), (iii) protein change, (iv) condition(s), clinical significance (last reviewed). We sought train number instances emerging them....
Some recent artificial neural networks (ANNs) claim to model aspects of primate and human performance data. Their success in object recognition is, however, dependent on exploiting low-level features for solving visual tasks a way that humans do not. As result, out-of-distribution or adversarial input is often challenging ANNs. Humans instead learn abstract patterns are mostly unaffected by many extreme image distortions. We introduce set novel transforms inspired neurophysiological findings...
Imagine trying to track one particular fruitfly in a swarm of hundreds. Higher biological visual systems have evolved moving objects by relying on both appearance and motion features. We investigate if state-of-the-art deep neural networks for tracking are capable the same. For this, we introduce PathTracker, synthetic challenge that asks human observers machines target object midst identical-looking "distractor" objects. While humans effortlessly learn PathTracker generalize systematic...
ABSTRACT Cereals are the key contributors to global food security. Genes involved in uptake (transport), assimilation and utilization of macro- micro-nutrients responsible for their content grain straw. Although many cereal genomic databases available, currently there is no cohesive web-resource manually curated nutrient use efficiency (NtUE) related genes QTLs, etc. In this study, we present a containing information on NtUE genes/QTLs corresponding available microRNAs some these four major...
Reinforcement Learning has recently surfaced as a very powerful tool to solve complex problems in the domain of board games, wherein an agent is generally required learn strategies and moves based on its own experiences rewards received. While RL outperformed existing state-of-the-art methods used for playing simple video games popular it yet demonstrate capability ancient games. Here, we one such problem, where train our agents using different namely Monte Carlo, Qlearning Expected Sarsa...
Detecting tremors is challenging for both humans and machines. Infants exposed to opioids during pregnancy often show signs symptoms of withdrawal after birth, which are easy miss with the human eye. The constellation clinical features, termed as Neonatal Abstinence Syndrome (NAS), include tremors, seizures, irritability, etc. current standard care uses Finnegan Scoring System (FNASS), based on subjective evaluations. Monitoring FNASS requires highly skilled nursing staff, making continuous...
There are genes whose function remains obscure as they may not have similarities to known regions in the genome. Such 'unknown' constituting Open Reading Frames (ORF) that remain epigenome termed orphan and proteins encoded by them but having no experimental evidence of translation 'Hypothetical Proteins' (HPs).We enhanced our former database Hypothetical Proteins (HP) human (HypoDB) with added annotation, application programming interfaces descriptive features. The hosts 1000+ manually...
Recent neural network architectures have claimed to explain data from the human visual cortex. Their demonstrated performance is however still limited by dependence on exploiting low-level features for solving tasks. This strategy limits their in case of out-of-distribution/adversarial data. Humans, meanwhile learn abstract concepts and are mostly unaffected even extreme image distortions. Humans networks employ strikingly different strategies solve To probe this, we introduce a novel set...
Delayed cancer detection is one of the common causes poor prognosis in case many cancers including oral cavity. Despite improvement and development new efficient gene therapy treatments, very little has been done to algorithmically assess impedance these carcinomas. In this work, we attempt annotate viable attributes datasets for identification gingivobuccal (GBC). We further apply supervised unsupervised machine learning methods revealing key candidate GBC prognosis. Our work highlights...
Nearly all models for object tracking with artificial neural networks depend on appearance features extracted from a "backbone" architecture, designed recognition. Indeed, significant progress has been spurred by introducing backbones that are better able to discriminate objects their appearance. However, extensive neurophysiology and psychophysics evidence suggests biological visual systems track using both motion features. Here, we introduce $\textit{PathTracker}$, challenge inspired...
Adversarial attacks can affect the object recognition capabilities of machines in wild. These often result from spurious correlations between input and class labels, are prone to memorization large networks. While networks expected do automated feature selection, it is not effective at scale object. Humans, however, able select minimum set features required form a robust representation an In this work, we show that finetuning any pretrained off-the-shelf network with Extreme Image...
ABSTRACT All annotated genes were once hypothetical or uncharacterized. Keeping this as an epilogue, we have enhanced our former database of proteins (HP) in human (HypoDB) with added annotation, application programming interfaces and descriptive features. The hosts 1000+ manually curated records the known ‘unknown’ regions genome. new updated version HypoDB functionalities (Blast, Match) is freely accessible at http://www.bioclues.org/hypo2 .