Integrative Personalized Omics Profiling Next Steps: Detection and Classification of Deviations from Wellness

Precision medicine is an emergent interdisciplinary field that utilizes state-of-the art technologies to detect changes in individual wellness, diagnose, treat, and possibly prevent diseases. Previous studies, by us and others, have focused on establishing the feasibility of personalized precision medicine, by measuring multiple time signals in individuals (generalized omics). These signals typically include medically relevant measures, such as clinical tests, blood panels, gene expression, and data from personal health devices. However, algorithmic development and sophisticated methods to analyze, classify, and biologically annotate such omics are still underdeveloped. In our proposed investigation, we will focus on the essential next steps to implement such integrative Personal Omics Profiling (iPOP) on individual astronauts, and detect and classify adverse medical events through the analysis of generalized omics signals.

Objectives: Our objectives are to: (i) Generate new computational methodology to establish the baseline healthy characteristics in individual personnel, and detect baseline deviations that predict the onset of any adverse medical event. (ii) Classify departures from individualized wellness according to collective biological/physiological signal characteristics and behavior. We will annotate the classes based on medical relevance, for instance by associating groups of signals to a specific disease. Our methodology will compare medical baseline deviations both within and across individuals.

Methods: To achieve our objectives we plan to develop sophisticated computational methods in statistical time series analysis and dynamic network theory to characterize and categorize different biological/physiological signals from individuals. We will develop and implement new algorithms that establish an individual's health baseline through the integration of thousands of signals, and that can detect deviations from a person's own baseline. The time series will be analyzed in terms of their underlying frequencies and limitations due to missing data or uneven sampling will be addressed. We will create a geometric representation of the signals, including networks that represent biological interactions between the omics that will allow us to both detect anomalies, but also infer the medical relevance of abnormal medical events.

Significance to solicitation's objectives: Our proposed investigation is completely aligned with Topic 1 of the solicitation, to develop Predictive algorithms of health, behavior, and medical events. The frameworks and methods of this investigation are designed to be directly applied to the monitoring of spaceflight personnel. Deliverables of this project include: (i) novel computational methodology, (ii) open source software packages (including a standalone analysis tool), and (iii) a report assessing the predictive utility of different data types and tissue samples that may be used in flight (including saliva).

Significance to NASA's interests and programs: Personalized monitoring of astronauts is already being performed, and through implementing our novel methods we will be able to improve NASA's ability to detect baseline abnormalities in flight, intervene with corrective measures, and maintain individualized wellness for personnel, which is mission critical.

We have continued to develop new algorithms that can identify changes in health, and potential medical events. All our methods are extensible and can be used to incorporate any dynamic data that can be measured during spaceflight - including different sampling rates (this includes molecular as well as device data that can be quantified). Furthermore, our approach can be implemented on Earth as well using personnel medical monitors to detect medical events.

The frameworks and methods of this investigation were primarily designed to be applied to the monitoring of individual spaceflight personnel. We have extended our trend detection algorithms (implemented in the open source Python package PyIOmica, and updated in Year 2 with new functionality and python compatibility) that address multiple challenges faced in analyzing signal data from human subjects and in-mission personnel. Specifically our trend detection algorithms and community detection methods:

(i) integrate thousands of time series per person,

(ii) apply to multiple signals of varying origin (e.g., transcripts versus metabolites versus radiation measurements),

(iii) address different time frames over which measurements may be taken (e.g., seconds versus hourly or daily measurements),

(iv) use many or few time points in the analysis,

(v) address missing data,

(vi) address uneven sampling of signals (typically the case for in-flight or in-the-field measurements),

(vii) integrate multiple layers of information (from different measurements, using different monitors),

(viii) provide utilities for visualization and facilitating the interpretation of data by non-specialists (including pathway annotations for gene-based or metabolite-based measurements),

(ix) provide visibility graphs methodology to detect and visualize temporal communities corresponding to changing physiological states in time series signals.

The methods can be used to analyze individualized changes (per subject) compared to their own baseline, and also provide data that can be used to compare between-subject changes to identify common differential changes corresponding to potential medical events that show common responses across subjects.

Deliverables of this project include: (i) novel computational methodology, (ii) open source software packages (including a standalone analysis tool), and (iii) assessing the predictive utility of different data types and tissue samples that may be used in flight (including saliva).

We have implemented the analysis to data for validations including:

(i) the evaluation of saliva multi-omics for non-invasive diagnostics and evaluation of predictive information per omics-data type analyzed in an individual subject to immune activation via vaccination.

(ii) the evaluation of individual subjects subject to immune activation (using retrospective data from prediabetic cohort analyisis).

(iii) the evaluation of retrospective astronaut data, from long-duration missions to assess the effects of spaceflight, and detect events that may be associated to medical in-flight events.

Significance to NASA's interests and programs: Personalized monitoring of astronauts is already being performed, and through implementing our novel methods we will be able to improve NASA's ability to detect baseline abnormalities in flight, intervene early with corrective measures, and maintain individualized wellness for personnel, which is mission critical. Furthermore, saliva omics offers non-invasive diagnostics, which can potentially be extended to spaceflight monitoring in long-duration space missions.

This is early/mid-stage (TRL 2) — the most realistic path in is NASA SBIR/STTR, which funds small businesses and research institutions to develop technology aligned with NASA's needs (equity-free, phased funding). Check whether a current SBIR/STTR solicitation topic overlaps with this project's technology area, or contact the project directly (above) to ask.

None of these are guaranteed paths for this specific project — TechPort itself doesn't have an "apply" button. Reaching out to the contact(s) above with a specific question is usually the fastest way to find out what's actually open.