A5059324690- Green IT and Sustainability
- Bioinformatics and Genomic Networks
- Health, Environment, Cognitive Aging
- Functional Brain Connectivity Studies
- Gene expression and cancer classification
- Genetic Associations and Epidemiology
- Cloud Computing and Resource Management
- Scientific Computing and Data Management
- Microbial Metabolic Engineering and Bioproduction
- Computational Drug Discovery Methods
- Energy, Environment, and Transportation Policies
- Environmental Impact and Sustainability
- Smart Grid Energy Management
- Air Quality Monitoring and Forecasting
- demographic modeling and climate adaptation
- EEG and Brain-Computer Interfaces
- Reproductive Physiology in Livestock
- Municipal Solid Waste Management
- Protein Structure and Dynamics
- Optical Imaging and Spectroscopy Techniques
- IoT and Edge/Fog Computing
- Machine Learning in Materials Science
- Mobile Crowdsensing and Crowdsourcing
- Genetics, Bioinformatics, and Biomedical Research
- Advanced Data Storage Technologies
University of Cambridge
2019-2024
British Heart Foundation
2020-2024
Health Data Research UK
2020-2024
Software Sustainability Institute
2024
Genomics (United Kingdom)
2023
Roland Hill (United Kingdom)
2019
University College London
2019
The Royal Free Hospital
2019
Abstract Climate change is profoundly affecting nearly all aspects of life on earth, including human societies, economies, and health. Various activities are responsible for significant greenhouse gas (GHG) emissions, data centers other sources large‐scale computation. Although many important scientific milestones achieved thanks to the development high‐performance computing, resultant environmental impact underappreciated. In this work, a methodological framework estimate carbon footprint...
Bioinformatic research relies on large-scale computational infrastructures which have a nonzero carbon footprint but so far, no study has quantified the environmental costs of bioinformatic tools and commonly run analyses. In this work, we estimate bioinformatics (in kilograms CO2 equivalent units, kgCO2e) using freely available Green Algorithms calculator (www.green-algorithms.org, last accessed 2022). We assessed 1) approaches in genome-wide association studies (GWAS), RNA sequencing,...
Machine learning and deep models have become essential in the recent fast development of artificial intelligence many sectors society. It is now widely acknowledge that these has an environmental cost been analyzed studies. Several online software tools developed to track energy consumption while training machine models. In this paper, we propose a comprehensive introduction comparison for AI practitioners wishing start estimating impact their work. We review specific vocabulary, technical...
Protein-protein interactions (PPIs) are essential to understanding biological pathways as well their roles in development and disease. Computational tools, based on classic machine learning, have been successful at predicting PPIs silico, but the lack of consistent reliable frameworks for this task has led network models that difficult compare discrepancies between algorithms remain unexplained.
Abstract Given that scientific practices contribute to the climate crisis, scientists should reflect on planetary impact of their work. Research computing can have a substantial carbon footprint in cases where researchers employ computationally expensive processes with large amounts data. Analysis human neuroimaging data, such as Magnetic Resonance Imaging brain scans, is one case. Here, we consider ten ways which those who conduct research reduce computing, by making adjustments studies are...
Given that scientific practices contribute to the climate crisis, scientists should reflect on planetary impact of their work. Research computing can have a substantial carbon footprint in cases where researchers employ computationally expensive processes with large amounts data. Analysis human neuroimaging data, such as Magnetic Resonance Imaging brain scans, is one case. Here, we consider ten ways which those who conduct research reduce computing, by making adjustments studies are planned,...
Abstract Computationally expensive data processing in neuroimaging research places demands on energy consumption—and the resulting carbon emissions contribute to climate crisis. We measured footprint of functional magnetic resonance imaging (fMRI) preprocessing tool fMRIPrep, testing effect varying parameters estimated and performance. Performance was quantified using (a) statistical individual‐level task activation regions interest (b) mean smoothness preprocessed data. Eight variants...
Abstract Genetically predicted levels of multi-omic traits can uncover the molecular underpinnings common phenotypes in a highly efficient manner. Here, we utilised large cohort (INTERVAL; N=50,000 participants) with extensive data for plasma proteomics (SomaScan, N=3,175; Olink, N=4,822), metabolomics (Metabolon HD4, N=8,153), serum (Nightingale, N=37,359), and whole blood Illumina RNA sequencing (N=4,136). We used machine learning to train genetic scores 17,227 traits, including 10,521...
Abstract Bioinformatic research relies on large-scale computational infrastructures which have a non-zero carbon footprint. So far, no study has quantified the environmental costs of bioinformatic tools and commonly run analyses. In this study, we estimate footprint (in kilograms CO 2 equivalent units, kgCO e) using freely available Green Algorithms calculator ( www.green-algorithms.org ). We assess (i) approaches in genome-wide association studies (GWAS), RNA sequencing, genome assembly,...
Abstract Protein-protein interactions (PPIs) are essential to understanding biological pathways as well their roles in development and disease. Computational tools, based on classic machine learning, have been successful at predicting PPIs silico , but the lack of consistent reliable frameworks for this task has led network models that difficult compare discrepancies between algorithms remain unexplained. To better understand underlying inference mechanisms underpin these models, we designed...
Computationally expensive data processing in neuroimaging research places demands on energy consumption - the resulting carbon emissions contribute to climate crisis. We measured footprint of fMRI preprocessing tool fMRIPrep, testing effect varying parameters estimated and performance. Performance was quantified using (a) statistical individual-level task activation regions interest (b) mean smoothness preprocessed data. Eight variants fMRIPrep were run with 257 participants who had...
We compared the carbon emissions of preprocessing and statistical analysis fMRI data in software packages FSL, SPM, fMRIPrep. Carbon for fMRIPrep were 30x larger than those 23x SPM. also scientific performance each package, reflected by sensitivity to activation. Overall, demonstrated slightly superior both FSL with outperforming However, this pattern varied brain region. Researchers analysing can use information inform their choice considering footprint processing alongside usability...
Abstract Scientists, empowered by huge amounts of computing power, storage and memory are making world changing discoveries — including to help combat climate change. Loïc Lannelongue, a PhD student at the University Cambridge, explores how high performance itself can lighten its carbon contribution.
Climate change is profoundly affecting nearly all aspects of life on earth, including human societies, economies and health. Various activities are responsible for significant greenhouse gas emissions, data centres other sources large-scale computation. Although many important scientific milestones have been achieved thanks to the development high-performance computing, resultant environmental impact has underappreciated. In this paper, we present a methodological framework estimate carbon...
Machine learning and deep models have become essential in the recent fast development of artificial intelligence many sectors society. It is now widely acknowledge that these has an environmental cost been analyzed studies. Several online software tools developed to track energy consumption while training machine models. In this paper, we propose a comprehensive introduction comparison for AI practitioners wishing start estimating impact their work. We review specific vocabulary, technical...