Antimicrobial resistance (AMR) in leprosy (“Hansen’s disease”) is an emerging global threat to leprosy control. Diagnosing AMR is a challenging process with animal model experiments that require specialized molecular laboratories to culture Mycobacterium leprae (M. leprae). Gene mutations within drug resistance determining regions of the genome of M. leprae are also known to be associated with phenotypic resistance to anti-leprosy drugs. We present HARP (Hansen’s Disease Antimicrobial Resistance Profiles): an open access web database with computationally predicted impacts of all possible mutations in 3 drug target proteins. It enables clinicians and researchers in leprosy-endemic countries to determine the mutation severity on AMR outcomes.
Blundell’s laboratory (at the Department of Biochemistry, University of Cambridge, England) includes a multidisciplinary team (computational and experimental arms) of senior research associates, post doctoral research scientists, PhD students and project interns in drug discovery programs for Cancer and infectious diseases (e.g. TB, leprosy, cystic fibrosis). In 2016, American Leprosy Missions (ALM) initiated a project: “Structure Guided Fragment-based drug discovery for leprosy”. We are a team of 1 Sr. research associate (SRA), 2 postdoctoral research associates, 1 PhD student and 2 project fellows. Dr. Sundeep Vedithi (SRA) led the HARP initiative as a part of the drug discovery program bringing together departmental TB and Cystic Fibrosis bioinformatics researchers. Team members worked on collecting information from various open access (as well as in-house developed) mutation impact predictors. For some of the tools, collaborators at the University of Melbourne helped collect the data from their servers. Data collected was managed on in-house data servers and made accessible to all. Currently the data is regularly updated by the research group in Cambridge and managed on third party servers. ALM maintains the server.
HARP was developed from Oct. 2018–Sep. 2020. Project Goals:
Neglected tropical diseases affect one billion people annually. Of these, 200,000 new leprosy cases are reported – a steady rate for 20 years. If individuals fail to respond to multidrug therapy (MDT) comprising of Dapsone, Rifampicin and Ofloxacin (a second line therapy) due to emerging AMR, there are no effective alternative strategies to combat leprosy.
About 3,000 relapse leprosy cases are reported annually; most are resistant to one or all MDT drugs. With no global system for active surveillance of leprosy AMR, diagnosis is mainly by clinical examination and molecular lab analysis. Once mutations are detected, there is a need for rigorous bioinformatics analysis to decipher the association between mutations and resistance outcomes. This process is outsourced to genomic companies, and the timeline is usually 1-2 weeks. With HARP, researchers can conduct this analysis in a few minutes at no cost.
HARP has an Open Knowledge Foundation license for public reuse of data without any restrictions for usage/reproduction. It has an interactive web design and also has supplementary tables along with published manuscript (PMID: 33304465). Protein structure as protein-data-bank files and tables can be freely downloaded using various editable formats. Users can interactively view the protein structure files on the Mol* interface and make appropriate analysis. They can securely upload their own data and conduct analysis. They can extract data from the database using command line on Linux (promoting access even with slow internet connectivity and basic computational infrastructure). Researchers can therefore easily reproduce the structural bioinformatics and reuse the data for their own research.
HARP is a demonstration of best practices for universities/advanced scientific centers that are addressing critical issues: they can create online platforms to share important findings and encourage local groups (especially those in resource limited settings) to carry out innovative on-the-ground research.
HARP was developed using open access python Flask framework. We used readily available tools. Data was collected from the online and in-house web servers for predicting impact of mutations on resistance outcomes. Data was generated using python scripts to submit queries to these web servers and extract data. The database was set up using Postgresql and Psycopg-2 relational database development architecture. Front end web design was performed using Flask HTML. Pandas dataframes enable users to browse data tables and download information. Linux command line and Python based API algorithms enable users to download all the data from the web and conduct their own analysis or reuse the information.
HTML blocks can be accessed with “inspect” command in the right mouse menu on the webpage where links to JavaScripts can be identified (all are left for open access/editing to customize visualizations on local version of Mol* for further use with proteins of interests). Links to all database development web servers have been provided on the help pages of the website.
Data sharing and reuse is an important channel to propagate science and information enabling reproducibility and implementation of findings: HARP promotes this. We track the usage of this web resource using google analytics platform (1,500+ unique page views last year). Data is captured on the google studio reports and analyzed to understand the level of usage and application of findings through citations on google scholar.
We strongly encourage best practices for sharing research data that contributes to collaborative research and/or supports innovative programs addressing critical health issues. The biocomputing group where HARP was developed has launched many data sharing portals/web servers which supported research programs globally addressing infectious diseases. Unless connected to patents and/or governed by proprietary rights, research data should be shared on open and easily accessible portals encouraging reproducibility and testing. These approaches promote translation of research findings to practice. One of our goals is to support researchers in leprosy endemic countries to easily diagnose a case of leprosy AMR. Although the data is purely on the research end, the learnings can be implemented in the diagnostic context contributing to leprosy control and eradication. We have overcome barriers by producing resources that address critical gaps in access and technological limitations. We would strongly urge researchers (producing non-proprietary biomedical data) to use appropriate existing online resources created by NCBI (National Centre for Biotechnogy Information) or create their own web resources to share with researchers worldwide.
