Data Science-Precision Health Training Program

Picture of PI and trainees at NLM Conference.
From left to right: Bogdan Pasaniuc (PI), Nicola Churchill (trainee), Mara Pleasure (trainee), Aditya Pimplaskar (trainee), and Alex Bui (co-PI) at the NLM Conference in 2023.

The Institute for Precision Health and the Department of Computational Medicine is pleased to offer the Biomedical Data Science Training Program for Precision Health Equity.  The program is funded by the National Library of Medicine (T15LM013976) for five years, starting July 1, 2022.  Funding for both predoctoral trainees and postdoctoral trainees are available.

 
LEADERSHIP AND FACULTY | PREDOCTORAL PROGRAM | POSTDOCTORAL PROGRAM | WORKSHOPS | LIST OF FACULTY

The program combines didactic training in computational and statistical methods for large-scale analyses involving biomedical data (e.g., multi-omics; electronic health records, EHRs). Trainees will acquire breadth in biomedical informatics and data science, as well as targeted learning in specific topics around equity, bioethics, and precision health. A team science-based approach is taken throughout the learning experience. All trainees will have a mentorship team that co-supervises them through a research experience in precision health. The mentorship team will pair the trainee with at least one faculty with a computational background and at least one other with a clinical background in the relevant area of the project. The training program will foster a diverse, interdisciplinary environment for a new generation of scientists to learn to invent and develop the computational approaches and tools that deliver on the promise of precision health for everyone. Our program synthesizes elements across four foci, providing each student an understanding of core topics in each area:

Healthcare/clinical informatics. The intersection of clinical informatics and precision health involves the development, evaluation, and translation of methods using the electronic health record (EHR) for deep phenotyping; using artificial intelligence (AI)-based methods to individually-tailor clinical decision support; digital health platforms for communicating information to patients (e.g., return of results); as well as targeted types of analysis (e.g., radiomics, mHealth).

Translational bioinformatics. The genesis of modern precision medicine came with the Human Genome Project and newfound ability to analyze individual sequences. Precision health now embraces computational methods in an ever-widening (multi-)omics space; the development of polygenic risk scores to guide personalized decisions regarding diagnosis and treatment; and the connection between molecular and phenotyping presentation.

Clinical research informatics. The power of precision health stems from our ability to conduct analysis on big data, uncovering new patterns across high-dimensional datasets that mix observational data types over time. The computational complexity of these analyses drives new algorithms and biostatistical approaches, not only for improved scalability/efficiency but also to address bias detection and issues around reproducibility.

Public health informatics. Precision health and computational epidemiology methods are increasingly synergistic, providing more refined insights to public health agencies to formulate better policies (e.g., optimizing resource allocation); and an understanding of how data-driven analyses at the individual level can help to overcome health disparities and influence global health research.

For more information, contact Stacey Beggs, Program Manager in the Department of Computational Medicine: sbeggs@mednet.ucla.edu

LEADERSHIP AND FACULTY

Principal Investigators:

  • Bogdan Pasaniuc, Associate Professor, Departments of Pathology and Laboratory Medicine, Computational Medicine, and Human Genetics
  • Alex Bui, Professor, Department of Radiological Sciences; Director, Medical Informatics Program

Leadership Team: 

  • Wei Wang, Professor, Departments of Computer Science and Computational Medicine
  • Onyebuchi Arah, Professor, Department of Epidemiology
  • Dan Geschwind, Professor, Department of Neurology, Human Genetics, and Psychiatry; Director, Institute for Precision Health
  • Eleazar Eskin, Professor, Department of Computer Science; Chair, Department of Computational Medicine.

Internal Advisory Board: 

  • Clara Lajonchere, Deputy Director, Institute for Precision Health 
  • Steven Dubinett, Director, Clinical & Translational Science Institute; Interim Dean, David Geffen School of Medicine
  • Michael Steinberg, Director of Clinical Affairs, Jonsson Comprehensive Cancer Center
  • Greg Payne, Director, Graduate Programs in Biosciences
  • Jayathi Murthy, Dean, Henry Samueli School of Engineering and Applied Science 

Administrative Team:

  • Stacey Beggs, Program Director, Department of Computational Medicine
  • Gayane Hovhannisyan, Executive Assistant, Department of Computational Medicine

Participating Faculty

Alongside our leadership team, the program involves more than two dozen individuals who serve as research advisors for trainees, teach core and/or elective courses, and may participate in its operational committees. These faculty cover a broad spectrum of biomedical informatics and data science research and represent 14 primary home departments from the Schools of Medicine, Engineering & Applied Sciences, Public Health, and the College of Letters & Science, and multiple areas of computation, including artificial intelligence (machine learning, reinforcement learning, natural language processing); statistics and biostatistics; biomedical informatics (clinical, imaging, bioinformatics, public health); precision medicine; computational epidemiology; and clinical medicine.

PREDOCTORAL PROGRAM

Applicants Must Meet the Following Criteria:

1) Pursuing a Ph.D. in computer science, biological/health science, information sciences, mathematics, statistics, or another area of relevance. (Related graphic)

2) US citizen, non-citizen national, or permanent resident.  Non-citizen nationals are persons born in US outlying possessions.

3) Have not received more than 5 years of aggregate NRSA support at the predoctoral level (any combination of individual & institutional awards).

Application Package:

For positions starting July 1, 2023, predoctoral applicants must complete the following two steps by May 15, 2023:

1) Information Sheet: Please complete the google form and upload your application package, assembled as a single PDF, to the form.  The application package should include:

  • Applicant Statement: A general description of your proposed research project including aims, data point collection, and potential outcomes, highlight your research/education accomplishments, and provides background information relevant to your interest in this program. (1,500-word limit)
  • Current curriculum vitae (CV)
  • Your most recent academic transcript 

Note: Applicants who have already completed at least one year of the Ph.D. program, please enter the name of the faculty member who has agreed to serve as your mentor for the duration of the fellowship.  

2) Letters of recommendation:  The form will ask you to provide the names of two faculty members who have agreed to write a letter of recommendation to support your application.  One of them should be the UCLA professor who has agreed to serve as your faculty mentor (excluding new Ph.D. students). Please ask your professors to send the letter as an attachment to Gayane Hovhannisyan.

Description:

The predoctoral training program will provide foundational training in a specific discipline as well as in precision health. Trainees will be simultaneously part of the Training Program as well as a student in the Bioinformatics, Biomathematics, Computer Science, Epidemiology, or Medical Informatics Ph.D. programs. As such, they will receive the same amount of formal training as any of their peers in these Ph.D. programs, while also receiving specialized training in precision health. The training program includes courses (which may fulfill elective requirements in the Ph.D. program), additional formal educational activities such as workshops, and a set of informal activities: seminar series, journal clubs, and related activities. The student’s specific training plan will be designed by an advising team consisting of the trainee's faculty mentor and additional program faculty who have expertise related to the research interests and academic interests of the trainee. The trainees will all take a set of core courses together and participate in many of the program activities as a group. Participation in this T15 will not add to the overall matriculation time of its students.

The Program Directors will serve as first-year students’ advisors until the student chooses a faculty mentor. The PDs may recommend coursework to address identified deficiencies in prior training. By the end of the first year, the student will select a faculty mentor and a mentorship team that includes at least one mentor with a computational background and at least one other with a clinical background in the relevant area of research.  The training program will support a PhD student for up to three years; the students will be reappointed each year assuming they make adequate progress towards their degree.

Four core courses provide foundational training across the affiliated Ph.D. programs. (They are also core classes in the respective Ph.D. programs.)

  • Bioinformatics CM221 (Introduction to Bioinformatics)
  • Computer Science M226 (Machine Learning in Bioinformatics)
  • Bioengineering/Medical Informatics M227 (Medical Information Infrastructures and Internet Technologies)
  • Epidemiology/Biostatistics 203B (Introduction to Data Science)

Students in the program will also choose two electives to provide a deeper understanding of a specific area of interest within precision health. Focus areas may include statistical genetics, algorithmic bias detection, radio-genomics, health equity, societal genetics, population health, and genomic risk prediction. 

Trainees will take Rigor and Reproducibility (MOL BIO 235), Reproducibility in Epidemiologic Research (EPI 207) or a comparable course or training approved by the director.

Finally, all trainees will complete the Collaborative Institutional Training Initiative (CITI) training offered through the campus Office of the Human Research Protection Program (OHRRP), as well as an in-person course that includes extensive content on RCR. Trainees may choose one of three courses: 

  • BIOMATH M261 (Responsible Conduct of Research Involving Humans)
  • HPM 225A (Health Services Research Design)
  • MIMG C234 (Ethics and Accountability in Biomedical Research)
  • EPI 207 (Reproducibility in Epidemiologic Research)

POSTDOCTORAL PROGRAM

Applicants Must Meet the Following Criteria:

1) Holds a doctoral degree (M.D. or Ph.D.) in computer science, biological/health science, information sciences, mathematics, statistics, or another area of relevance at the time of appointment.  A pending degree is acceptable to apply; degree confirmation is required at the time of official appointment.

2) US citizen, non-citizen national, or permanent resident.   Non-citizen nationals are persons born in US outlying possessions.

3) Have not received more than 3 years of aggregate NRSA support at the postdoctoral level (any combination of individual & institutional awards). 

Postdoctoral Application Package:

Applicants for postdoctoral training must already be considered for a postdoctoral position at UCLA in any department and have identified the faculty member who will serve as their mentor (see list of faculty). It is the applicant’s responsibility to obtain the mentor’s approval to conduct their proposed work as a member of their UCLA laboratory.

For positions starting July 1, 2023, postdoctoral applicants must complete the following two steps by May 15, 2023:

1) Information Sheet: Please complete the google form and upload your application package, assembled as a single PDF, to the form.  The application package should include:

  • Applicant Statement: A general description of your proposed research project including aims, data point collection, and potential outcomes, highlights your research/education accomplishments, and provides background information relevant to your interest in this program. (1,500-word limit)
  • Current curriculum vitae (CV)
  • Your most recent academic transcript 

Note: Please enter the name of the faculty member who has agreed to serve as your mentor for the duration of the fellowship.  

2) Letters of recommendation:  The form will ask you to provide the names of two faculty members who have agreed to write a letter of recommendation to support your application. One of them should be the UCLA professor who has agreed to serve as your faculty mentor.  Please ask your professors to send the letter as an attachment to Gayane Hovhannisyan.

Description:

Postdoctoral trainees are required to complete assigned coursework/training based on their backgrounds and scholarship goals, conduct a mentored research project related to precision health, complete a course in responsible conduct of research (RCR) and training in scientific reproducibility and help lead at least two workshops.

The faculty mentor, along with two additional faculty will be chosen will provide tailored mentorship for all postdoctoral trainees. The trainee’s mentoring team will consist of the faculty mentor and two additional faculty: one mentor with a computational background and one with a clinical background in the relevant area of research.  The mentoring team will work closely with the postdoc trainee to design the training plan. There are two recommended paths:

Training for postdoctoral trainees with advanced informatics/data science degrees. As it is unlikely that a trainee will already be knowledgeable in all four areas covered by this Training Program and precision health, the mentorship team will identify courses within the predoctoral core curriculum and electives (see predoctoral program) to provide the requisite background and exposure.

Training for postdoctoral trainees without advanced informatics/data science degrees. A postdoctoral trainee without a background in biomedical informatics or data science will complete the Biomedical Informatics track of the MSCR during their training period. The Master of Science in Clinical Research is a 2-year program that provides rigorous training in the methodology and techniques utilized in biomedical informatics. The curriculum provides didactic course work including the computational and statistical methods. These postdocs will complete the four predoctoral program core courses, which count toward the MSCR, as well as the following courses: 

  • Biomathematics 170A (Introductory Biomathematics for Medical Investigators)
  • Biomathematics M260C (Methodology in Clinical Research III: ObservationalStudies)
  • Biomathematics 266A (Applied Regression Analysis in Medical Sciences)
  • Biomathematics 266B(Advanced Biostatistics)
  • Bioengineering 220 (Introduction to Medical Informatics)
  • Bioengineering M226 (Medical Knowledge Representation)

The required thesis for the MSCR will be integrated with the mentored research project that is part of the training program.

***

Trainees will take Rigor and Reproducibility (MOL BIO 235), Reproducibility in Epidemiologic Research (EPI 207) or a comparable course or training approved by the director.

Finally, all trainees will complete the Collaborative Institutional Training Initiative (CITI) training offered through the campus Office of the Human Research Protection Program (OHRRP), as well as an in-person course that includes extensive content on RCR. Trainees may choose one of three courses: 

  • BIOMATH M261 (Responsible Conduct of Research Involving Humans)
  • HPM 225A (Health Services Research Design)
  • MIMG C234 (Ethics and Accountability in Biomedical Research)

WORKSHOPS AND OTHER ACTIVITIES

Participation in workshops provide further exposure and depth across areas of biomedical informatics/data science and precision health. These workshops highlight pragmatic issues in working with data and the translational challenges. All trainees will take the UCLA Health Data Discovery Resource training (see below). In addition, predoctoral trainees are required to select another two workshops, while postdocs will be expected to help lead two workshops.

UCLA Health DDR training. The UCLA Data Discovery Resource is a de-identified extract of the UCLA’s clinical electronic health record, covering data from 2013 on over 1.5M patients. The Institute for Quantitative and Computational Biosciences, in collaboration with UCLA Health System, offers a workshop on how to access the data and the computational resources in an integrated cloud computing environment.

QC Bio Collaboratory Workshops. QCB offers over two dozen workshops related to data analysis and programming, many of which are directly relevant to this training program. Workshops cover the skills and knowledge that will help the students successfully carry out their research. Workshop lectures are accompanied by problems and exercises performed by the students in the classroom. 

Computational Genomics Summer Institute (CGSI). Computational Medicine hosts a month-long NIH-funded research education program for mathematical and computational scientists, sequencing technology developers, and biologists who use genomic technologies for research applications. The first week focuses on computational biosciences broadly, the second week provides tutorial presentations, mentoring, and small-group discussions for a smaller group of trainees, and the third week provides a series of talks on human disease and precision health. It has been enormously successful at creating a computational genomics community and at training and mentoring trainees who are beginning their research careers. 

Mentor Training. All trainees will acquire mentor training through UCLA’s Graduate Programs in Biosciences. Predoctoral trainees will enroll the 10-week spring course entitled "Entering Mentoring Training" (MOL BIO 300).  Postdoctoral trainees will take the six-hour Postdoc Research Mentor Training.

NLM Informatics Training Conference. All trainees in the program will attend this annual conference held in June. The grant provides travel funding.

Other activities of the Training Program include a journal club, career and professional development workshops, and a seminar series.