Biology subjects

Bouazoune, K.

Publications and source records attributed to Bouazoune, K..

3 recordsLinked to original sources

bioRxiv2024

Optimizing biomedical information retrieval with a keyword frequency-driven Prompt Enhancement Strategy

BackgroundMining the enormous pool of biomedical literature to extract accurate responses and relevant references is a daunting challenge, made more difficult by the domains interdisciplinary nature, specialized jargon, and continuous evolution. Early natural language processing (NLP) approaches faced considerable hurdles in comprehending the nuances of natural language which frequently led to false answers. However, the discovery of transformer models has immensely boosted the cause by enabling researchers to create larger and more potent language models known as large language models (LLMs). These LLMs have opened up new possibilities for enhancing question-answering (QA) tasks. Even with these technological advances, the current LLM-based solutions for querying specialized domains like biology and biomedicine still have issues in generating up to date responses while preventing "hallucination" or the generation of plausible but factually incorrect responses. ResultsThis paper focuses on prompt enhancement using retrieval-augmented architecture as a strategy to guide LLMs towards generating more meaningful responses for biomedical question-answering tasks. We evaluated two prompt enhancement approaches using GPT-3 and GPT-4, examining their effectiveness in retrieving relevant information from complex landscape of biomedical literature. Our proposed approach leverages explicit signals in user queries to extract meaningful contexts from a vast pool of information, addressing the shortcomings of traditional text embedding-based prompt enhancement methods. We developed a QA bot WeiseEule (https://github.com/wasimaftab/WeiseEule-LocalHost) utilizing these prompt enhancement methods and allowing for comparative analysis of their performances. ConclusionsOur findings highlight the importance of prompt enhancement methods that utilize explicit signals in users query over traditional text embedding based counterparts to improve LLM-generated responses in specialized domains such as biology and biomedicine. By providing users complete control over the information that goes into the LLM, our approach tackles some of the major drawbacks of existing web-based chatbots and LLM-based QA systems including hallucinations and the generation of irrelevant or outdated responses.

bioinformatics↗

bioRxiv2020

CHD7 interacts with the nucleosome acidic patch for its efficient activity via its N-terminal region

Chromodomain-Helicase DNA binding protein 7 (CHD7) is an ATP dependent chromatin remodeler involved in maintaining open chromatin structure. Mutations of CHD7 gene causes multiple developmental disorders, notably CHARGE syndrome. However, there is not much known about the molecular mechanism by which CHD7 remodels nucleosomes. Here, we performed integrative biophysical analysis on CHD7 chromatin remodeler using crosslinking-mass spectrometry (XL-MS), cryo-Electron Microscopy (cryo-EM) and single-molecule Forster Resonance Energy Transfer (smFRET). We uncover that N-terminal to the Chromodomain (N-CRD) interacts with nucleosome. Importantly, this region is required for efficient ATPase stimulation and nucleosome remodeling activity of CHD7. The cryo-EM analysis on the N-CRD_Chromodomain bound to nucleosome reveals that the N-CRD interacts with the acidic patch of nucleosome. Furthermore, smFRET analysis shows the mutations in the N-CRD result in slow or highly-fluctuating remodeling activity. Collectively, our results uncover the functional importance of a previously unidentified N-terminal region in CHD7 and implicate that the multiple domains in chromatin remodelers are involved in regulating their activities.

biochemistry↗

bioRxiv2020

Overarching control of autophagy and DNA damage response by CHD6 revealed by modeling a rare human pathology

Members of the chromodomain-helicase-DNA binding (CHD) protein family are chromatin remodelers critically implicated in human pathologies, with CHD6 being one of its least studied members. Here, we discovered a de novo CHD6 missense mutation in a patient clinically presenting the rare Hallermann-Streiff syndrome (HSS). We used genome editing to generate isogenic iPSC lines and model HSS in relevant cell types. We show that CHD6 binds a cohort of autophagy and stress response genes across cell types. The HSS-mutation affects CHD6 protein folding and impairs its ability to recruit co-factors in response to DNA damage or autophagy stimulation. This leads to an accumulation of DNA damage burden and to senescence-like phenotypes. By combining genomics and functional assays, we describe for the first time a molecular mechanism for the chromatin control of autophagic flux and genotoxic stress surveillance that applies broadly to human cell types and explains HSS onset.

cell biology↗