FOA Title: 
Notice of Joint DMS/NLM Initiative on Generalizable Data Science Methods for Biomedical Research.
Grant Type: 
NOT-LM-19-001
Primary IC: 
NLM
Release Date: 
Oct 10 2018
Expiration Date: 
N/A
AC Source: 
N/A
Purpose: 
Notice Joint DMS/NLM Initiative Generalizable Data Science Methods Biomedical Research. Notice Number: NOT-LM-19-001 Key Dates Release Date:October 10, 2018 Related Announcements None Issued National Library Medicine NLM) Purpose Significant advances technology coupled decreasing costs associated data collection storage resulted unprecedented access vast amounts health- disease-related data.  Biomedical data includes genomics data next-generation sequencing, data different imaging modalities, real-time static data wearable electronics, personal mobile devices, environmental sensors, observational health data, clinical data hospitals, insurance, electronic medical records including personal health records. The National Library Medicine NLM) the Division Mathematical Sciences the Directorate Mathematical Physical Sciences DMS) the National Science Foundation NSF) recognize need support research develop innovative transformative mathematical statistical approaches address important data-driven biomedical health challenges. The goal this interagency program the development generalizable frameworks combining first principles, science-driven models structural, spatial temporal behaviors innovative analytic, mathematical, computational, statistical approaches can portray fuller, nuanced picture a person's health the underlying processes. program designed foster inter- multi-disciplinary collaborations.  Collaborative efforts bring together researchers the biomedical/health the mathematical/statistical sciences communities a requirement this program must convincingly demonstrated the proposal. particular interest new collaborative efforts involving mathematicians, statisticians, biomedical scientists, clinicians aimed blending first principles, science-based models innovative data-driven machine learning approaches solve important biomedical problems.  While research be motivated a specific application dataset, development methods are generalizable broadly applicable preferred encouraged. of important application areas currently supported the National Library Medicine include following: Finding biomarkers support effective treatment through integration genetic Electronic Health Records EHR) data; Understanding epigenetic effects human health; Extracting analyzing information EHR data; Understanding interactions genotype phenotype humans linking human sensor data genomic data using dbGaP; Protecting confidentiality personal health information; Mining heterogeneous data sets e.g. clinical environmental). list not intended be exhaustive exclusive.  However, proposals should clearly discuss the intended new collaborations address biomedical challenge describe use publicly-available biomedical datasets validate proposed models methodology. NIH datasets related the research themes listed above include Clinicaltrials.gov ( https://clinicaltrials.gov/ ) Image data repository - Clinical Center ( https://www.nih.gov/news-events/news-releases/nih-clinical-center-releas... Model Organism Databases ( https://www.genome.gov/10001837/model-organism-databases/  ) RefSeq NCBI ( https://www.ncbi.nlm.nih.gov/refseq/ ) Human Connectome Project ( https://www.humanconnectome.org) Adolescent Brain Cognitive Development ABCD) Data Repository ( https://data-archive.nimh.nih.gov/abcd ) Applicants expected list specific datasets will used the proposed research demonstrate they access these datasets.  Data Management Plan should describe plans make data available researchers these data not the public domain. program designed promote development sophisticated mathematical, statistical, computational models methods address biomedical data science challenges, such    Modeling integration heterogeneous data different sources; Incorporation synthetic data address bias a data set; Development methods handle spatio-temporal dependencies missingness; Causal Inference Machine Learning; Model validation, uncertainty quantification, evaluation, reproducibility, metrics FAIR findable, accessible, interoperable reusable); Natural Language Processing approaches that address combinations structured/unstructured text. Application submission through National Science Foundation via solicitation NSF-19-500 ( https://www.nsf.gov/pubs/2019/nsf19500/nsf19500.htm?org=NSF).Specific information concerning application review process available ( https://www.nsf.gov/pubs/2019/nsf19500/nsf19500.htm?org=NSF). those applications are being considered potential funding NLM, PD s PI s be required submit ir application s an NIH-approved format. PD s PIs invited submit NIH receive further information submission procedures. applicant not allowed increase proposed total budget change scientific content the application the submission the NIH . results the first level scientific review be presented NLM Board Regents the second level review. NLM make final funding determinations issue Notices Awards successful applicants. NLM DMS anticipate making 8 10 awards totaling to 4 million, fiscal year 2019.  is expected each award be between 200,000 300,000 total costs) per year durations up 3 years. Inquiries Please direct inquiries to: Jane Ye, PhD National Library Medicine NLM) Telephone: 301-594-4882 Email: yej@mail.nih.gov