PSB 2027

PSB 2027 Workshop

Next‑Generation Biocomputing Doctoral Training in the Era of AI

Lead organizer: Graciela Gonzalez‑Hernandez, PhD (Cedars‑Sinai) –  Professor and Vice Chair of Computational Biomedicine; Director of Health AI graduate programs embedded in a major health system. Long‑standing leader in doctoral program design and graduate curriculum innovation across biomedical informatics and Health AI, including major roles in designing PhD programs at the University of Pennsylvania and Arizona State University. Cedars‑Sinai Medical Center endorses her involvement in organizing this workshop.

Co‑organizer: Steven E. Brenner, PhD (UC Berkeley) – Professor of Computational Genomics in the Department of Plant and Microbial Biology and member of the Computational Biology graduate program. Long‑standing leader in computational biology education and doctoral‑level training, including extensive involvement in graduate program development, interdisciplinary training grants, and mentoring of PhD students at the interface of genomics, computation, and AI. The University of California, Berkeley endorses his involvement in organizing this workshop.

Invited speakers

  • Karin Verspoor, PhD (RMIT University) – Dean of the School of Computing Technologies; leader in international AI/NLP and digital‑health training initiatives.
  • Ben Raphael (Princeton University) – Professor Department of Computer Science, Princeton University, developed computational and mathematical methods for human and cancer genomics, including algorithms for identifying genomic alterations and tumor evolution.
  • John H. Holmes, PhD (University of Pennsylvania) – Senior biomedical informatics educator with extensive experience in doctoral training and curriculum innovation.

Invited speakers will also serve as panelists in the open discussion. Confirmed speakers have agreed to participate contingent on workshop acceptance; invited speakers are in active discussion and will be finalized upon notification.

Motivation and Goals

PhD programs in biocomputing disciplines (including health AI, biomedical informatics, computational biomedicine, and related areas) are under pressure to integrate modern AI/ML and LLM methods, strengthen interdisciplinary training, and prepare graduates for a widening spectrum of careers in academia, industry, and health systems, all while maintaining rigorous methodological and socio‑technical foundations. Recent opinion pieces and discussions have highlighted concrete challenges: difficulty integrating AI/ML into already dense curricula, tension between traditional informatics content and new technical demands, and a widening skills gap between academic training and real‑world AI practice.

Specifically, these questions have surfaced repeatedly in recent PSB meetings, where participants have raised concerns about whether existing doctoral training in biomedical informatics and related biocomputing disciplines is keeping pace with AI‑driven changes in research practice. In the 2026 PSB discussions in particular, AI‑related educational issues—ranging from how to teach with and about generative models to how to assess AI‑assisted work—were debated vigorously, underscoring the need for a focused, community‑wide conversation in the form of this workshop. Leaders and trainees from established and emerging doctoral programs attending PSB will have in this workshop a space to:

  • Contrast structural and curricular approaches to PhD training in Health AI and biomedical informatics.
  • Surface key design tensions and tradeoffs in the AI/LLM era, including how generative AI is changing research practice, mentoring, and assessment.
  • Identify practical principles and collaborative opportunities around resources, assessment models, practicum structures, self‑curated learning, and cross‑institutional activities.

Technical area and format

The workshop focuses on doctoral education across biocomputing disciplines (health AI, biomedical informatics, computational biomedicine, and computational biology) in the era of AI/LLMs. It combines six short talks on distinct program models, a guided panel discussion on design tensions, and a practicum‑style closing block with two focused talks (assessment/AI use and vibe coding) plus a structured whole‑room discussion that produces concrete experiments participants can implement in their own programs.

Target Participants

  • PhD students and postdocs who want to shape their own training paths in Health AI and biomedical informatics, including design of self‑curated learning experiences.
  • Faculty who supervise PhD students or teach graduate‑level AI, informatics, software development or data‑science courses.
  • Program directors and graduate program leaders (including invited speakers) who wish to share and compare models.
  • Health‑system and industry AI leaders who co‑mentor PhD trainees or host doctoral projects.

The session is designed so that PhD students, postdocs, and individual faculty leave with at least one concrete change they can pursue in their own mentoring relationships, lab practices, graduate‑level courses, or self‑directed learning, even if they do not control an entire program. We will capture a short list of concrete “experiments for the next year” and identify volunteers to coordinate follow‑up activities (such as a shared resource repository or a perspective/consensus paper on doctoral training in Health AI and biomedical informatics in the era of pervasive AI and vibe coding).