Clinical Research Informatician My principal
Description: Clinical Research Informatician My principal scientific focus is on finding ways to use observational data to answer medical questions with less bias. Since causal knowledge is a prerequisite to ascertain whether a statistical association
Related Topics
Download Presentation
"Clinical Research Informatician My principal" is the property of its rightful owner. Permission is granted to download and print the materials on this website for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Presentation Transcript
slide1. Clinical Research Informatician
My principal scientific focus is on finding ways to use observational data to answer medical questions with less bias. Since causal knowledge is a prerequisite to ascertain whether a statistical association between an exposure and an outcome is causal, my approach is to use computable knowledge mined from the literature and other sources to build graphical model applications. I develop methods that are both data-driven and knowledge-driven that learn data generating processes. My work which has appeared in such prestigious venues as the Journal of Biomedical Informatics and AMIA proceedings, is strongly interdisciplinary, combining elements of epidemiology, graph theory, and text mining. I am always looking for new collaborators in these fields. Currently, I am conducting research from both causal inference (quantifying causal estimands) and causal discovery (structure learning) perspectives in the following application areas:
the detection of harms in EHR data for drug safety and comparative effectiveness research; and
High-throughput drug repurposing for Alzheimer’s disease (and other degenerative diseases) from signals in observational data. Scott Alexander Malec, PhD
NLM K99/R00-funded Postdoc Fellow
Contact: Email: sam413@pitt.edu
Phone: (412) 330-7082
LinkedIn: /in/scottalexandermalec
ORCID: https://orcid.org/0000-0003-1696-1781 https://docs.google.com/presentation/d/1mY7Tbz-JiP3hXRquzzlEdj3qvyO_zBatAp5lXqXkkp0/edit#slide=id.p<br>
slide2. Literature-informed Pharmacovigilance Key Tools
Knowledge: SemMedDB
Causal inference: TMLE, CEM
Codebase: currently in R, postgres
GitHub: /kingfish777/causalSemantics Project Details
Funding: NLM T 15 (8/1/18-7/31/21),
NLM K99/R00 (8/1/2021-7/31/25)
[ 1K99LM013367-01A1 ]
Collaborators: PITT (SoM, Pharm), UW-Seattle,
UTH-Houston https://doi.org/10.1016/j.jbi.2021.103719<br>
slide3. Drug Repurposing for Alzheimer’s Disease Key Tools
Knowledge: SemMedDB, OBO
Knowledge representation: Pheknowlator Knowledge Graph
Machine readers: SemRep (NIH/NLM); REACH, EIDOS (DARPA)
Causal inference: TMLE, CEM
Codebase: python, R, postgres
GitHub: /dbmi-pitt/alz_lbgcm & /kingfish777/semiramis Project Details
Funding: Pitt MOMENTUM (2020), NLM T 15 (8/1/18-7/31/21),
NLM K99/R00 (8/1/2021-7/31/25)
[ 1K99LM013367-01A1 ]
Collaborators: PITT (SoM, ISP, SPH), U Colorado, UW-Seattle Third-factor phenotype variables linking depression with Alzheimer’s disease<br>
slide4. Combining Epidemiology Best Practices with Data Science<br>
My principal scientific focus is on finding ways to use observational data to answer medical questions with less bias. Since causal knowledge is a prerequisite to ascertain whether a statistical association between an exposure and an outcome is causal, my approach is to use computable knowledge mined from the literature and other sources to build graphical model applications. I develop methods that are both data-driven and knowledge-driven that learn data generating processes. My work which has appeared in such prestigious venues as the Journal of Biomedical Informatics and AMIA proceedings, is strongly interdisciplinary, combining elements of epidemiology, graph theory, and text mining. I am always looking for new collaborators in these fields. Currently, I am conducting research from both causal inference (quantifying causal estimands) and causal discovery (structure learning) perspectives in the following application areas:
the detection of harms in EHR data for drug safety and comparative effectiveness research; and
High-throughput drug repurposing for Alzheimer’s disease (and other degenerative diseases) from signals in observational data. Scott Alexander Malec, PhD
NLM K99/R00-funded Postdoc Fellow
Contact: Email: sam413@pitt.edu
Phone: (412) 330-7082
LinkedIn: /in/scottalexandermalec
ORCID: https://orcid.org/0000-0003-1696-1781 https://docs.google.com/presentation/d/1mY7Tbz-JiP3hXRquzzlEdj3qvyO_zBatAp5lXqXkkp0/edit#slide=id.p<br>
slide2. Literature-informed Pharmacovigilance Key Tools
Knowledge: SemMedDB
Causal inference: TMLE, CEM
Codebase: currently in R, postgres
GitHub: /kingfish777/causalSemantics Project Details
Funding: NLM T 15 (8/1/18-7/31/21),
NLM K99/R00 (8/1/2021-7/31/25)
[ 1K99LM013367-01A1 ]
Collaborators: PITT (SoM, Pharm), UW-Seattle,
UTH-Houston https://doi.org/10.1016/j.jbi.2021.103719<br>
slide3. Drug Repurposing for Alzheimer’s Disease Key Tools
Knowledge: SemMedDB, OBO
Knowledge representation: Pheknowlator Knowledge Graph
Machine readers: SemRep (NIH/NLM); REACH, EIDOS (DARPA)
Causal inference: TMLE, CEM
Codebase: python, R, postgres
GitHub: /dbmi-pitt/alz_lbgcm & /kingfish777/semiramis Project Details
Funding: Pitt MOMENTUM (2020), NLM T 15 (8/1/18-7/31/21),
NLM K99/R00 (8/1/2021-7/31/25)
[ 1K99LM013367-01A1 ]
Collaborators: PITT (SoM, ISP, SPH), U Colorado, UW-Seattle Third-factor phenotype variables linking depression with Alzheimer’s disease<br>
slide4. Combining Epidemiology Best Practices with Data Science<br>