Checking sign-in…

Skip to handbook

INTERNAL HANDBOOK · MEMBERS ONLY

LABS Guidebook

Onboarding, IT, Mercury Wi‑Fi, computing, meetings, reproducibility, templates, time off, and offboarding—reference for lab members.

Read handbook Access

Access

Member access

This handbook is for signed-in lab members. Use the table of contents to jump to onboarding, IT, and policies. If you need a new invite (for example after a rejected application), contact your PI or lab admin—registration is not linked from here.

Overall

LABS mission and research focus

LABS focuses on developing and applying AI/ML methodologies to multi-organ, multi-omics data to understand human aging and disease. We emphasize the following aspects:

  • AI/ML is the engine
  • Clinical neuroscience, implications, and impacts on patient care are the goal
  • Reproducibility and objectivity are the basics

Working time

LABS requires full-time members, such as postdocs, RAs, and PhD students, to work in person for all normal workdays (9 AM to 5 PM). The PI does not micromanage and expects members to effectively and wisely manage work/life balance.

All interns must come to the lab 1–2 days a week, especially if they have been assigned a specific project and are granted computational access, and can discuss suitable days/times on Discord.

That said, we recognize that individual circumstances may differ, and alternative arrangements can be discussed directly with the PI.

Onboarding

Onboarding checklist

Admin contact

Radiology

Required accounts (e.g., GitHub, institutional VPN, etc.)

Set up an account for the cluster, GitHub, VPN, etc.

Discord channels

Ask the PI to invite you to join the LABS’ Discord channel.

Safety and compliance training (IRB, HIPAA, etc., if applicable)

Follow the university’s policies and complete all required courses. If you have doubts, contact your PI for advice.

IT

Getting on Mercury Wi‑Fi

Must be on the Columbia VPN: download Cisco AnyConnect Secure Client connect plug-in — ssl.cpmc.columbia.edu — and then log in using Duo push.

Go to this link to register your device: CUMC ClearPass onboarding. It’ll register your device — details of the steps are here: CUMC Wi‑Fi / Mercury.

Now you can log in to Mercury using your UNI login and password.

Printing in the lab

To get access to the printer in the lab, you must reach out to IT: mm6310@cumc.columbia.edu.

Computing

Computing & learning resources

HPC

  • CUBIC
  • Empire AI
  • Insomnia

Dataset

UK Biobank training

Learning materials

Learning links

Code

Code & reproducibility

All papers before submission should be cleaned for reproducibility under the LABS GitHub account: https://github.com/LABS-lab.

  • Lab GitHub or GitLab structure.
  • Code review and pull request expectations.
  • Use of version control, documentation standards (e.g., READMEs).
  • Best practices for notebooks and reproducible scripts.

Our research underscores the principles of open science and reproducibility. Please have a GitHub account and follow LABS’ official account: https://github.com/LABS-lab. All source code of your project will be made publicly available before your paper is submitted to journals and conferences. You will remain the developer of the code, but the copyright will belong to the lab.

Meetings

Meetings & communication

Weekly lab meeting schedule

We have a weekly meeting on Fridays from 9:30–11:30, including all lab members, to discuss project advancement.

Weekly journal clubs

Our journal seminar is scheduled for every Friday afternoon at 2:00 PM. If you plan to present on a Friday, please share the papers in Discord that you intend to discuss with the schedule several days in advance, specifically by the Monday of that week, so everyone has the opportunity to read the papers beforehand. Click here.

1-on-1 meeting

Personal meetings will be scheduled based on people’s advancement on specific projects.

Communications

Discord: We use Discord for all scientific discussions, idea and paper sharing, but don’t share any raw data on Discord. For all data, share it through the HPC internally in the lab (/cbica/projects/MULTI/dropbox).

Email: For any formal communications, please communicate via email; again, no individual raw data is permitted in emails.

GitHub: For collaborative projects like software development, you need to use git and GitHub, especially since we are a computationally heavy lab.

Responsible use of AI

Responsible use of AI

LLMs have inevitably been involved in our research and life, and we need to ensure a fair and ethical use of these AI tools.

Coding: Lab members may leverage generative AI to generate code snippets based on task descriptions, as well as to assist with code debugging, editing, and auto-completion. As the human agents at the core of this iterative loop, you assume full responsibility for reviewing all AI-generated content to ensure a thorough understanding of the code, task requirements, and all associated details.

Writing: Lab members may use AI assistance for report drafting; however, the core concepts, original insights, and intellectual content must originate from the authors themselves. The uploading of entire documents, with the expectation that AI will complete the entirety of the writing, is strictly prohibited. Please bear in mind: in this new era of scientific research, you remain at the very center of the process and must, therefore, assume full responsibility for all your research endeavors. For example, we encourage lab members whose native language is not English to use LLMs to refine their writing and perform grammar checks; however, please be sure to carefully review all content.

Template

Template

Presentation: Template related to conference presentation: Click here.

AI conference LaTeX: e.g., NeurIPS, CVPR, ICML, and ICLR.

Nature journal presubmission inquiry: Click here; the PI prefers using Word for edits, but LaTeX is fine too.

Citation management: Zotero

Figures

We encourage the use of R for all figure generation because it is publication-ready and more elegant than Python.

InkScape: Use this software to edit your SVG vector image (X for raster image) so that it can be shared and edited by your colleagues.

Personal time off

Personal time off

All full-time employees in LABS (postdoc and PhD students) should follow the University-wide policy regarding personal time off (except official holidays and weekends).

If you wish to request personal days off, please email the PI to confirm the specific dates.

Here are specific policies for Radiology-based employees (e.g., postdoc) and PhD students:

At the end of each calendar year, please return an Excel file (click here) to the PI, denoting your total time-off days throughout the year.

Leave the lab

Leave the lab

Academic moves are common, particularly for early-career researchers, and we aim to make each transition as smooth as possible. Please follow the procedures below.

First, complete all required paperwork in accordance with Columbia University and Radiology policies (for postdoc and RA, etc.), or the policies of your respective BME or CS department (PhD students).

For all researchers who actually work with the data and project in the lab, these steps must be completed by the last working day or before account deactivation, unless approved by the PI.

Follow CU and Radiology or your respective BME or CS department policy for all related paperwork.

Besides, specific to LABS:

  • Return all lab-owned equipment, including laptops, hard drives, and any other devices.
  • Push all code related to your work in the lab to the LABS GitHub account before submitting any manuscripts.
  • Code README and usage instructions
  • Environment or dependency details
  • Project status summary
  • Locations of data, scripts, and results
  • Unresolved issues or next steps
  • Securely remove all data related to the lab’s research from your possession if you will not have continuing collaboration with the lab. For example: no copies on personal laptops, external drives, personal cloud storage, email, or private GitHub repositories.
  • If your project is still ongoing at the time of your departure, please coordinate with the PI to determine how the work will continue, whether through your involvement or by transferring responsibilities to other lab members. Authorship decisions will be based on actual contributions by the PI. In general, you are welcome to continue working on the project based on your own plan and willingness, and additional responsibility (e.g., revision and communication with other lab members) is needed.
  • Unpublished ideas, analyses, code, protocols, and results developed in the lab may be confidential and cannot be shared externally without permission; coordinate with the PI.

Please send an email stating all these items to the PI before your last day in the lab.