Courses

BIOS 667 · Applied Longitudinal Data Analysis

GEEs, mixed models, joint modeling, and missing-data tactics grounded in public health and medical case studies. Course re-implemented from scratch with lectures embedding live R code examples and illustrations, complemented by AI-friendly assignments that emphasize reproducible workflows.

BIOS 735 · Statistical Computing

Machine learning, git/GitHub, simulation, optimization, and package development for translational teams. Project-based with open-source code reviews. Course site

Reproducible Oncology Analytics Workshop

Git, Quarto notebooks, and genomic-study case studies tailored for LCCC T32 trainees.

Mentoring & Student Outcomes

Five completed PhD advisees (co-advised with Drs. Ibrahim, Love, Kosorok).

Academic placements:

  • Pedro Baldoni (2020) — Assistant Professor, University of Pittsburgh Biostatistics
  • Hillary Heiling (2023) — Senior Biostatistician, Dana-Farber Cancer Institute

Industry placements:

  • Euphy Wu (2024) — Biostatistician, Precision Genomics
  • David Lim (2022) — Scientist, GSK
  • Scott Van Buren (2020) — Scientist, GSK

Training approach: Independent research aims, embedded clinician collaborations, weekly manuscript meetings, and reproducible software deliverables. I help students prepare for the careers they want — whether in academia, industry, or government — and tailor their projects to best position them for their goals. Students typically complete 2-4 first-author publications during their PhD program.

For Prospective PhD Students

Information about training structure, technical expectations, and lab environment for students considering applying.

Coding & software development

Required: R programming (or strong willingness to learn intensively in Year 1). Students should be comfortable with data manipulation, visualization, and basic statistical modeling.

Developed during training: Git/GitHub workflows, R package development, reproducible research practices (Quarto/RMarkdown), high-performance computing. Python, Julia, or C++ for specialized projects as needed.

Software contribution: Most students develop new packages as part of dissertation work. CRAN/Bioconductor submission expected for methods projects.

Mentoring cadence

Individual meetings: Weekly 30-60 min during active project phases; biweekly during writing/revision periods.

Lab meetings: Weekly group manuscript meetings where students present works-in-progress for feedback.

Mentoring approach: Hands-on in Year 1-2 (weekly code review, collaborative writing); increasingly independent in Years 3-5 (student-driven agenda, advisor feedback). Students drive their own research direction by dissertation proposal stage.

Publication expectations

Dissertation projects: Students are first author on all dissertation aims (typically 2-3 papers). Advisor is senior author unless methods are co-developed equally.

Collaborative projects: Authorship determined by contribution; students are first author when leading analysis and writing. Typical timeline: 6-12 months from data to submission for collaborative work; 12-24 months for methods papers.

Typical outcome: Students graduate with 3-4 first-author publications (mix of methods and collaborative) plus 2-4 co-authorships.

Financial support model

Typical funding: Students are typically 100% GRA-funded through lab research grants.

Grant applications: Students are encouraged to apply for F31 (predoctoral fellowship) or departmental training grants. Advisor provides proposal feedback and budget support.

Conference travel: Funding available for 1-2 conferences per year (JSM, ENAR) when presenting work.

Disease areas & flexibility

Current focus: Breast and pancreatic cancer genomics, adaptive trial design, ctDNA modeling, tumor-stroma interactions.

Methodological breadth: Statistical methods are often disease-agnostic (missing data, high-dimensional modeling, Bayesian adaptive designs). Students interested in immunotherapy, radiomics, or other cancer types are welcome.

Student-driven exploration: Students have flexibility to propose new application areas if methodologically aligned with lab expertise.

Work culture & resources

Collaboration: Embedded in UNC Lineberger with regular contact with cancer researchers. Students often contribute and support computational and statistical needs of ongoing Lineberger studies.

Computational resources: Access to Longleaf HPC cluster, lab GitHub organization, shared codebases for common workflows.

Work expectations: Flexible hours; no expectation of evening/weekend work except near deadlines.

Apply through the UNC Biostatistics PhD program. Prospective students interested in the lab can email naim@unc.edu with questions about research directions, lab culture, or training philosophy. Current UNC Biostatistics PhD trainees: email directly to discuss joining the lab or participating in collaborative projects.

Last updated: March 2026