The AI/ML AWG aspires to scientific excellence and participation is strictly on a voluntary basis. The NASA Open Science Data Repositories are responsible for collecting and making available space-relevant multi-omics and phenomics datasets. The datasets span a broad range of biological levels involving data from tissues, organs, whole organisms, physiology, and behavior covering the full gamut of life (e.g. microbes, plants, mammals, invertebrates, …). The feedback from the AI/ML AWG will aid in optimizing the Open Science Data Repositories’ data-sharing systems to enable dataset generation and training models along with various data science-AI approaches. The AI/ML AWG joins five other Open Science AWGs (the GeneLab Animal, Microbes, Multi-omics, and Plant AWGs, and the ALSDA AWG). Because the ‘-omics’ and molecular-level data within the GeneLab repository are often from the same studies and subjects found in the ALSDA data, AI/ML methods can be deployed to discover and characterize relationships across heterogeneous data types (e.g. microscopic and omics measurements from the same animals). This overall effort will maximize the scientific return on investment from underused spaceflight and space-relevant data. The AI/ML AWG aspires to yield:
User feedback on AI/ML-readiness of GeneLab and ALSDA Repositories. Feedback will be sought from AWG members on the user experiences regarding data readiness for dataset curation, annotation, accessibility, and reuse.
Science data management. From origination of the space science experiments to data collection to data reusage, standards and requirements are necessary to effectively manage the data workflow. The members of this AWG will provide guidance on optimal ways to curate and disseminate high dimensional datasets and pre-trained models for shared use.
Community standards for assay, sample, and study design metadata. Consensus will be sought on standards for assay, sample, and study design metadata which must be collected from PIs to ensure data reusability. Additionally, this goal also involves data format standards to ensure data is not only re-usable, but also suitable for machine learning applications.
Modeling and risk assessment. With the deployment of novel AI/ML techniques such as explainable AI, another goal is to provide tools to model and/or predict the response of various life forms to space stressors. Such tools may prove useful to translate data into human risk for spaceflight and precision space health.
Scientific communication of analyses and results. The AWG is expected to communicate the science results and applications of GeneLab and ALSDA data re-use. Such communications could take the form of self-published white papers, conference presentations, or peer-review publications. These communications will also serve to catalyze the generation of new multidisciplinary and collaborative avenues for research that can increase the value and impact of the data within the GeneLabs and ALSDA repositories.
AWG participants will determine how the AI/ML AWG best reaches these goals. Optimally, the AWG should hold regular virtual or in-person meetings to organize work and review progress. The schedule for delivering the above products is set by the Chair (or Co-Chairs) in consultation with GeneLabs, ALSDA, and the other AWG participants.
Each AI/ML AWG meeting must adhere to the following criteria: (1) it must have at least one participant who possesses biological domain area expertise; (2) it must have at least one participant who possesses machine learning expertise; (3) it must have at least one representative from GeneLab and/or ALSDA. AWG participants are chosen based on a candidate’s ability to provide the scientific or data analysis insight necessary to achieve AWG goals. AWG participation is entirely voluntary, and participants will not be compensated.
Chairs - The AWG will have a Chair (or Co-Chairs) selected by AWG participants. The Chair or Co-Chairs will lead the AWG and will draft a schedule, along with reviewing progress of the AWG towards achieving goals. Chairs will serve until they decide to step down, at which time GeneLab and/or the AWG participants will decide on and appoint a new Chair or Co-Chairs.
GeneLab and ALSDA Representatives - The AWG will have a designated representative from GeneLab as well as ALSDA to help ensure outcomes of the work meet the criteria above as well as support the work of the AWG. Representatives from GeneLab and ALSDA can also serve as a Chair or Co-Chair of the AI/ML AWG.
Coordinators – The following Coordinators will be selected on a volunteer basis by the AWG. Multiple coordinating functions may be performed by a single Coordinator.
Analysis Coordinator - The analysis coordinator tracks the progress of individual analyses, ensuring that analyses are done in a consistent and reproducible manner, sets deadlines for completion of individual analyses, and collects and organizes interim and final analysis results.
Communication Coordinator – Assembles information for newsletter or manuscript sections, ensures consistency in writing and figure styles, and sets goals for the completion of articles or newsletter.
Participation in GeneLab AWGs is open to all investigators (academic, government, private sector), and students (i.e. high-school, undergraduates, graduates, interns); however participants must be willing to:
Contribute to the goals of the AI/ML AWG,
Take part in group activities including periodic teleconferences. Participants are expected to actively contribute to the group goals (see above), and
Share AWG products with the AI/ML AWG participants for open collaboration.
An investigator that would like to take part in the AWG should provide a description of their expertise in the relevant area within machine learning to the AWG Chair/Co-Chairs. Requests for participation in AWGs will be reviewed by GeneLab and ALSDA, in consultation with the Chair/Co-Chairs, with an annual review of current membership.
By participating in the AI/ML AWG, investigators will join a group of scientific excellence. AWG participation may lead to co-authorship on science papers, as well as contributing towards operational and data-derived tools of great use for NASA and deep space mission support. The AI/ML AWG will help set data and analysis standards for the community and may develop relationships that could lead to additional collaborations.
Participants cannot use the AWG partnership as a means to claim any direct affiliation with NASA. Participants can, however, list partnerships with the AWGs on their CVs in the following manner, example below.
Active member (or Chair/Co-Chair) of the NASA Open Science AI/ML AWG since 2022. https://genelab.nasa.gov/awg/members
Invited speakers attending AWG meetings cannot claim AI/ML AWG or any other NASA partnerships as a result of their presentation. Any actions against this rule will result in an immediate termination of the member.
AWGs are strongly committed to ensuring that all meetings and interactions are conducted with the highest possible levels of integrity, inclusion, and professional respect. To this end, the AWGs are adopting the following Codes of Conduct for all members:
All participants are treated with respect and consideration, valuing a diversity of views and opinions
Be considerate, respectful, and collaborative. Communicate openly with respect for others, critiquing ideas rather than individuals
Anyone requested to stop unacceptable behavior is expected to comply immediately. Chairs, co-chairs, and GeneLab representatives may take any action deemed necessary and appropriate, including immediate removal from the meeting or the AWG
Do not interrupt others or talk over others
Be mindful of bias in all contexts
AWG chairs, co-chairs, and GeneLab representatives should step in immediately to address abusive or bullying behavior
Be respectful of all regardless of differences (professional or otherwise)
All members should work to create an environment free of harassment
Harassment, intimidation, or discrimination in any form will not be tolerated
Physical or verbal abuse of any participant
Examples of unacceptable behavior include, but are not limited to verbal comments related to gender, sexual orientation, disbility, physical apprearance, body size, race, religion, national origin, or threatening or stalking any participant.
If you are the subject of unacceptable behavior or have witnessed any such behavior, please notify chairs or GeneLab representatives immediately. Notification should be done by contacting the chairs and/or GeneLab representatives or by emailing your concern to: arc-dl-genelab-awg@mail.nasa.gov