Targeted Evaluation of Ionizing Radiation Exposure (TEI-REX) BAA.pdf

U.S. ARMY RESEARCH OFFICE

In partnership with

The Intelligence Advanced Research Projects Activity (IARPA)

BROAD AGENCY ANNOUNCEMENT FOR

Targeted Evaluation of Ionizing Radiation Exposure (TEI-REX)

W911NF-22-S-0002

Issued by:

U.S. Army Contracting Command-Aberdeen Proving Ground

Research Triangle Park Division

P.O. BOX 12211

Research Triangle Park, NC 27709-2211

I. OVERVIEW OF THE FUNDING OPPORTUNITY:

A. Required Overview Content

1. Federal Agency Name(s):

2. Funding Opportunity Title: Targeted Evaluation of Ionizing Radiation Exposure (TEI-REX)

3. Announcement Type

4. Research Opportunity Number: W911NF-22-S-0002

5. Catalog of Federal Domestic Assistance (CFDA) Number:

6. Response Dates:

B. Additional Overview Information

II. DETAILED INFORMATION ABOUT THE FUNDING OPPORTUNITY:

A Funding Opportunity Description 1 Program Summary 2 Scientific Premise for the Program 3 New Methods for Evaluating Lower-Dose Ionizing Radiation Exposure 4 Program Structure 5 Research Focus Areas by Phase 6 Recommended Team Expertise 7 Program Scope and Limitations 8 Program Test and Evaluation 9 Government Provided Equipment and Samples 10 Models, Software, and Application of Machine Learning/Artificial Intelligences Towards Biodosimetry 11 Program Waypoints, Milestones, and Metrics 12 Program Period of Performance, Timeline, and Deliverables 13 Meetings and Travel Requirements 14 Place of Performance

B. FEDERAL AWARD INFORMATION

C. Eligibility Information

1. Eligible Applicants:

2. Cost Sharing or Matching:

3. Federally Funded Research and Development Centers and University Affiliated Research Centers: ... 25

D. Application and Submission Information

1. Addresses to View Broad Agency Announcement

2. Content and Form of Application Submission Volume 1 – Technical & Management Proposal Volume 2 – Cost Proposal

3. Unique Entity Identifier and System for Award Management (SAM)

4. Submission Dates and Times:

5. Intergovernmental Review

6. Funding Restrictions:

7. Other Submission Requirements:

E. Proposal Evaluation Information

1. Criteria:

2. Review and Selection Process:

3. Recipient Qualifications

F. Award Administration Information

1. Award Notices:

2. Administrative and National Policy Requirements:

3. Reporting:

G. Agency Contacts

H. Other Information

1. Cooperative Agreement Proposals

2. Example of Technical Cover Sheet

3. Example of Academic Institution Acknowledgement Letter

4. Example of Technical SOW

5. Example of Team Organization Table

6. Example of Intellectual Rights Sheet

7. Example of Contract Deliverables Table

8. Example of Organizational Conflicts of Interest Certification Letter

9. Example of Three Chart Summary of the Proposal

10. Sample of the Research Data Management Plan

11. Cover Sheet – Cost Proposal

12. Example of Prime Contractor/Subcontract Cost Element Sheet for Volume 2: Cost Proposal

13. Example of Travel Costs Trip Breakdown Sheet

14. Glossary of Acronyms:

15. References

I. OVERVIEW OF THE FUNDING OPPORTUNITY:

A. Required Overview Content

1. Federal Agency Name(s):

U.S. Army Research Office

Issuing Acquisition Office:

U.S. Army Contracting Command-Aberdeen Proving Ground, Research Triangle Park Division (ACC-APG RTP Division)

2. Funding Opportunity Title: Targeted Evaluation of Ionizing Radiation Exposure (TEI-

REX)

3. Announcement Type Full Announcement

4. Research Opportunity Number: W911NF-22-S-0002

5. Catalog of Federal Domestic Assistance (CFDA) Number:

12.431 – Basic Scientific Research

6. Response Dates:

BAA release: 18 October 2021 Questions must be submitted by: 1 November 2021 5:00 PM EDT to TEI_REX- BAASubmission-2021@iarpa.gov Response to questions expected by 8 November 2021 Proposals due: 13 December 2021 @ 5:00 PM EDT Submitted to TEI_REX-BAASubmission-2021@iarpa.gov or via Grants.gov

B. Additional Overview Information

This Broad Agency Announcement (BAA) which sets forth research areas of interest to the Army Research Laboratory-Army Research Office (ARL-ARO) and the Intelligence Advanced Research Projects Activity (IARPA) is issued under paragraph 6.102(d)(2) of the Federal Acquisition Regulation (FAR), and 10 USC 2358 which provides for the competitive selection of basic research proposals. Proposals submitted in response to this BAA and selected for award are considered to be the result of full and open competition and in full compliance with the provision of Public Law 98-369, "The Competition in Contracting Act of 1984" and subsequent amendments.

The Department of Defense agencies involved in this program reserve the right to select for award;

all, some, or none of the proposals submitted in response to this announcement. The participating DoD agencies will provide no funding for direct reimbursement of proposal development costs.

Technical and cost proposals (or any other material) submitted in response to this BAA will not be returned. It is the policy of participating DoD agencies to treat all proposals as sensitive, competitive information and to disclose their contents only for the purposes of evaluation.

II. DETAILED INFORMATION ABOUT THE FUNDING

OPPORTUNITY:

A Funding Opportunity Description

1 Program Summary The objective of the Targeted Evaluation of Ionizing Radiation Exposure (TEI-REX) program is to establish novel biodosimetry approaches enabling improved quantification of lower-dose ionizing radiation exposures (<0.75 Gray)1 from samples that can be collected and/or tested minimally or non-invasively (hereafter known as ‘TEI-REX samples’), while also expanding quantitative and qualitative knowledge of the exposure environment. To accomplish this, the TEI- REX program will discover, characterize, and model biomarkers associated with a variety of ionizing radiation events, especially lower-dose exposures, from TEI-REX samples. The new capabilities developed under TEI-REX align with United States Government’s mission objectives ranging from investigation of exposure events to ensuring compliance with established dosimetry protocols.

Current biodosimetry approaches, including the gold standard dicentric chromosome assay (DCA), are effective for determining a higher-dose radiation exposure, but suffer from multiple limitations.

These constraints include: the need for invasively collected sample(s), such as blood; requiring multiple collections of the sample; a limited period for which a first sample must be collected post-exposure for an accurate prediction of dose exposure; a dependence on transient markers to calculate exposure dose, resulting in a limited period the test is effective following an exposure;

and wide standard deviations of dose calculations at lower-dose exposures. The TEI-REX program aims to establish and characterize novel biomarkers, which can overcome many of the limitations that current biodosimetry approaches do not address.

Recently published research has demonstrated that biomarkers associated with ionizing radiation exposure can be detected across numerous biological targets including proteins, peptides, metabolites, and lipids2. While recent research has focused on basic targets in simple models, they showcase the potential for expanding biodosimetry techniques to novel tissue and sample types, while overcoming limitations associated with current biodosimetry techniques. As these earlier efforts focused on high dose exposures and/or high dose rates ex vivo, TEI-REX is advancing this research by focusing on more complex sample targets in more complex (e.g., in vitro and in vivo) environments at lower dose exposures, with the hypothesis that irradiation of proteins and other biological targets will demonstrate a minimal- or no-threshold, non-linear sensitivity response to the exposure. This hypothesis should be considered distinct, but potentially overlapping, from the

1 One Gray (Gy) is the international system of units (SI) equivalent of 100 rads, which is equal to an absorbed dose of 1 Joule/kilogram.

2 (Benjamin B Minkoff, 2019; Jelena Tamuliene, 2020; Merriline M. Satyamitra, 2020; Elisabeth Vicente, 2020;

William Blakely, 2010; Younghyun Lee, 2018; Changran Geng, 2020) linear, no-threshold model used by the EPA to assess the clinical risk associated with lower-dose radiation exposure3.

2 Scientific Premise for the Program The concept of the TEI-REX program relies upon the law for the conservation of energy. Ionizing radiation contains energy, which is transferred to the organic and non-organic elements with which it interacts. In biological systems, as cells and intercellular spaces are predominately comprised of water, much of this linear energy transfer (LET) results in the generation of reactive species (e.g., reactive oxygen species (ROS) and reactive nitrogen species (RNS)), but some energy may directly transfer to biological materials, such as: DNA, proteins, metabolites or lipids (Figure 1). LET or reactive species may cause changes, often in the form of damage, to various components of the cell4. Most biodosimetry approaches rely upon calculating damage done to DNA or the downstream effects of irradiative damage. TEI-REX is researching the effects done to the other biological components, specifically those which are long lasting and directly attributable to the initial ionizing insult.

Lower doses of ionizing radiation will result in fewer molecular changes compared to higher-dose exposures, making detection even more difficult as lower-dose exposures approach ambient background, but the changes will still be present. Researchers recently demonstrated, when using free amino acids and 3-residue peptides, a consistent order of reactivities exists across amino acids when exposed to higher doses of radiation5. Other researchers have also shown that glutamine, for example, has unique electron impact fragmentation patterns based upon variable and high dose exposures6. Finally, additional researchers are in the early stages of characterizing irradiated proteins to demonstrate the potential for a biologically informed dosimeter7.

3 (United States Environmental Protection Agency, 2021) 4 (Keisz, Bansal, Qian, Zhao, & Furdui, 2014) 5 (Benjamin B Minkoff, 2019) 6 (Jelena Tamuliene, 2020) 7 (Changran Geng, 2020)

Figure 1: Overarching TEI-REX concept with notional signatures and detection methods.

3 New Methods for Evaluating Lower-Dose Ionizing Radiation Exposure The TEI-REX program seeks proposals for new methods of quantifying lower-dose (<0.75 Gray) ionizing radiation exposure to an organism, shifting away from methods which are dependent upon invasive sampling and more traditional DNA, RNA, and expression-based signatures. The objectives are to discover robust biomarkers associated with TEI-REX samples, develop methods for enabling the detection of these signatures, and develop computational models to interpret these signatures in relation to the exposure event. Ideally, these biomarkers will be conserved across organisms and/or sample types enabling improved extensibility. Realistically, the program recognizes the significant challenge this research objective represents and that initial steps likely will focus on a limited number of optimal model organisms and sample targets. Offerors are encouraged to describe how extensible their biomarker(s) may be to other samples or models but must focus on how they will apply their proposed research against the primary objectives of this program instead of theorizing towards an ideal objective.

4 Program Structure The TEI-REX program is anticipated to be a 3.5-year (42 months) effort, comprised of three (3) Phases. All three (3) Phases are being solicited under this BAA. Phase 1 will be 18 months in duration, Phase 2 will be 12 months, and Phase 3 will be 12 months. Each phase will encompass two (2) main technical Focus Areas:

Focus Area 1 (FA1): Signatures and detection methodologies for characterizing ionizing radiation exposures, within 25 days of the first exposure event, from TEI-REX samples.

Focus Area 2 (FA2): Signatures and detection methodologies for characterizing ionizing radiation exposures, greater than 90 days from the last exposure event, from TEI-REX samples.

Exposure events should be considered acute or near-acute events (achieving final dose within minutes or hours). Multiple acute events may occur for individual models within the timeframe established under each Focus Area.

Offerors must propose to all three (3) Phases and both Focus Areas of the Program under this BAA. Proposals that submit to only 1 Focus Area or less than 3 Phases per Focus Area will be considered non-compliant.

The goal of Phase 1 is the successful prediction of higher-dose, 1 to 4 Gray, exposure and the timeline of exposure, days to months, from at least one (1) TEI-REX sample. Performers are expected to make these predictions by evaluating and modeling unique biomarkers associated with TEI-REX samples and using them to develop an effective biodosimetry model. These dose ranges were selected as the most likely to enable discovery of radiation induced biomarkers, without being suppressed by the noise associated with severe cellular damage and/or cell apoptosis. To accomplish this, performers are expected to establish a research pipeline for the discovery, characterization, and modeling of robust biomarker signatures, induced by ionizing radiation exposure. Offerors must propose research approaches that enable evaluation, aligning to TEI- REX metrics (Table 3) and the Test and Evaluation (T&E) approach (Section II, A, 8), of radiation dose exposures and the timeline of exposure through discovery of relevant biomarkers found in or on skin, hair, or other TEI-REX sample types (e.g., sweat, sebum, dermal interstitial fluid, hair follicles, etc.). Offerors must propose well-reasoned and supported research approaches for successfully identifying biomarkers from these higher-dose exposures in addition to modeling approaches that will meet Phase 1 metrics and T&E approaches.

The goal of Phase 2 is the successful prediction of lower-dose exposure, defined as ambient background8 to 0.75 Gray; the timeline of exposure, from days to months; the type of radiation (particulate, electromagnetic, and/or mixed), and; the dose rate of exposure (mGray/min) from at least two (2) TEI-REX samples. Performers are expected to leverage and optimize their efforts from Phase 1 to demonstrate successful analysis of samples at lower dose exposures, a greater variety of exposure time points, and a wider variety of samples/organisms while also improving upon overall confidence and accuracy of these predictions (Table 3). Offerors must propose improvements and optimizations for their research approach and describe why the approach has a high likelihood of success against the lower dose exposures and expanded Phase 2 metrics and T&E approaches.

Across both Phase 1 and 2 offerors may propose models distinct from those described under the T&E process in section II, A, 8. Any proposed approach must demonstrate technical strength

8 The natural radiation that is always present in the environment. It includes cosmic radiation which comes from the sun and stars, terrestrial radiation which comes from the Earth, and internal radiation which exists in all living things. The typical average individual exposure in the United States from natural background sources is about 300 millirems per year (https://www.nrc.gov/reading-rm/basic-ref/glossary/background-radiation.html).

and high likelihood of extensibility to T&E sample types. Offerors should not assume that blinded T&E activities will conform to specific performer unique model systems nor approaches, although adaptation to specific sample types may be feasible. The baseline for reference samples developed and provide by the program will be hair and skin from human-associated or mouse models.

The goal of Phase 3 is the successful evaluation of a range of realistic exposure scenarios across a range of doses, timelines, and organism types. Performers will be expected to integrate and expand the model systems developed under Phase 1 and 2 and apply their research efforts towards identifying the limits of the newly developed TEI-REX capabilities. This Phase will rely upon even more challenging metrics from Phase 2 (Table 3). Offerors will be expected to integrate, improve, and adapt their discovery, detection, modeling, and deductive approaches to evaluate at least 3 sample types. The most likely models and sample types will be discussed throughout Phase 2 and decisions will be released at the Phase 3 kickoff, offerors should plan for sample counts and testing timelines aligning with what is described further below.

5 Research Focus Areas by Phase Phase 1: Research the discovery, detection, and modeling of signatures associated with higher dose ionizing radiation exposures from samples that can be collected and/or tested minimally- or non-invasively.

Offerors are expected to describe how they will establish the exposure environment and the research efforts that support the discovery and characterization of biomarkers associated with higher dose exposures. Offerors must describe how their research approach aligns with TEI-REX objectives, to include planned radiation exposure environment, model system(s), sample type(s), analytical biomarker discovery and detection pipeline, and the biodosimetry modeling approach with strong and supported scientific reasoning. Offerors should account for the T&E approach described in Section II, 8 and adapt their research plan appropriately so they can evaluate samples or models distinct from the models and sample types used for internal research activities. Offerors must ensure their discovery and detection research proposal aligns with all program metrics in Table 3. Leveraging these systems to generate and discover appropriate biomarkers, offerors will be expected to model the biomarkers and associated data, enabling evaluation across a range of ionizing radiation doses for a given exposure timeline for each FA:

Phase 1/FA1: Research the discovery, detection, and modeling of signatures associated with higher dose, 1-4 Gray, ionizing radiation exposures to TEI-REX samples, within 25 days of a first exposure event.

Phase 1/FA2: Research the discovery, detection, and modeling of signatures associated with higher dose, 1-4 Gray, ionizing radiation exposures to TEI-REX samples, greater than 90 days from the last exposure event.

For both FA1 and FA2 in Phase 1, offerors will be responsible for describing their proposed model system(s), their research plan to include irradiation of the models(s) and evaluating appropriate TEI-REX sample(s), with the expectation to include biomarker research from at least one (1) sample type of either skin or hair. Offerors will develop their own protocols and systems for irradiation of models and samples. Offerors must describe their approach for ensuring lower-dose exposures across multiple radiation types in a highly consistent and quantitative manner; collection of samples, analysis of samples, discovery of biomarkers, and development of models from biomarkers enabling accurate prediction of exposure dose and dose timeline are all, at least, expected to be described in approach. Offerors will be expected to generate approximately 50 irradiated samples associated with their selected model system(s) and sample types to provide, in appropriate form and in a cost-effective manner, to T&E teams for evaluation of performer develop protocols. Offerors will be expected to evaluate T&E blinded samples, to include at least hair or skin samples, using the models developed throughout their research efforts (Figure 2). Offerors shall consider as deliverables: protocols, selected biomarker panels, irradiated samples, details on model systems, raw/processed/curated analytical data, and final outputs which includes reporting of alignment to program metrics (Table 3).

Phase 2: Research signatures and detection methods for characterizing lower dose ionizing radiation exposures from samples that can be collected and/or tested minimally- or non-invasively.

Offerors are expected to leverage the methods, models, platforms, and overall capabilities developed under Phase 1 to investigate biomarkers associated with lower-dose exposures. Offerors will be expected to improve upon metrics associated with biodosimetry modeling and biomarker detection as detailed under Table 3. Additionally, offerors are expected to research methods for applying biomarker analysis, from a single sample or multiple samples collected from a single model individual, towards assessing the type of radiation exposure (particulate, electromagnetic, or mixed) and the exposure dose-rate. The approaches for detecting biomarkers will be evaluated based upon the same metrics as the blinded T&E activities.

For both FA1 and FA2 in Phase 2, offerors will be responsible for optimizing their proposed model system(s), completing irradiation of the models(s), and evaluating appropriate TEI-REX samples from at least one (1) additional model beyond Phase 1 requirements, with the expectation to include biomarker research from at least one (1) sample type of either skin or hair. Submitted approaches should provide technical support describing potential extensibility of biodosimetry capabilities as such extensibility aligns to program objectives. Approaches which propose unsupported or technical unsound support for greater extensibility may be reviewed unfavorably.

Offerors will develop their own protocols for irradiation of models to include: collection of samples, analysis of samples and discovery of biomarkers, and development of models from biomarkers enabling accurate prediction of exposure dose and dose timeline. Offerors will be expected to generate approximately 50 irradiated samples associated with their selected model system(s) and sample types to provide, in appropriate form and in a cost-effective manner, to T&E teams for evaluation of performer develop protocols. Offerors will be expected to evaluate T&E blinded samples, to include at least hair or skin samples, using the models developed throughout their research efforts. Offerors shall consider as deliverables their protocols, selected biomarker panels, irradiated samples, details on model systems, raw/processed/curated analytical data, and final outputs which includes reporting of alignment to program metrics (Table 3). The lower dose exposure timeline for each Phase 2 FA is defined as follows:

Phase 2/FA1: Research the discovery, detection, and modeling of signatures associated with lower dose, background to 0.75 Gray, ionizing radiation exposures to TEI-REX samples, within 25 days of a first exposure event.

Phase 2/FA2: Research the discovery, detection, and modeling of signatures associated with lower dose, background to 0.75 Gray, ionizing radiation exposures to TEI-REX samples, greater than 90 days from the last exposure event.

Phase 3: Research with application of biomarker detection against samples generated to mirror realistic scenarios.

For both FA1 and FA2 in Phase 3, offerors are expected to optimize the capabilities developed under Phases 1 and 2, and integrate their protocols, biomarkers, and models, against a series of realistic and challenging model types, sample types, and/or confounders selected with direct input from program transition partners. Offerors should expect to test the viability and robustness of their biodosimetry pipeline against new model systems, to include the potential for larger mammals, insects, and plants. Additionally, a range of confounders, to include age, gender, UV exposure, chemical exposure, and the presence of natural antioxidants may be tested. Offerors must describe how their platform will adapt to these challenging sample types and propose how these factors may impact previously identified biomarkers. Evaluation of these samples will provide empirical evidence towards overall capability and extensibility, while also pushing offerors to research methods for challenging and niche use cases.

6 Recommended Team Expertise The research associated with the TEI-REX program is expected to incorporate a collection of diverse technical fields. Offerors are strongly encouraged to ensure all capabilities below are clearly identified with demonstrated expertise within their team. Expertise associated with an ideal TEI-REX program, not ordered by criticality, should include, but are not limited to:

1. Radiation biology

2. In vitro and in vivo models associated with radiation exposure

3. Analytical biochemistry

4. Biomarker discovery

5. Biodosimetry

6. Machine learning and Artificial Intelligence

7. Radiation dosimetry/health physics

8. Statistics

9. Program management

A recommended structure for reporting Team Expertise and organization is included in Section II, H as a Team Organization Table.

7 Program Scope and Limitations TEI-REX is focused on signatures enabling evaluation of ionizing radiation exposure, especially at lower-dose ranges, not in deriving or predicting clinical outcomes. While there may be future associations with clinical outcomes, this is not an objective of the program and offerors should maintain awareness of assumptions derived from clinical perspectives.

Proposals shall explicitly address all four (4) of the following:

I. Underlying theory: Proposals shall summarize their proposed models, samples, and methods for likelihood of robust biomarker detection. Detailed support reinforcing the technical approach should be included as referenced papers.

II. Research activities: Proposals shall describe the technical approach(es) being pursued to meet TEI-REX metrics and milestones for all three (3) phases.

III. Technical risks: Proposals shall identify technical risks and proposed mitigation strategies for each.

IV. Software or computational model development: Proposals shall describe the approach for developing software that enables effective and interpretable biodosimetry-based assessment of samples.

The following sample types, biomarkers, and confounders are out of scope for this program:

I. Any samples that cannot be collected and/or tested through minimally- or non-invasive means, unobtrusively. Samples which do not fit this category can still be utilized for initial research efforts, but final capabilities and associated biomarkers must align with these minimally- or non-invasive sample types in an offeror’s proposal. Types of collected samples that fall outside of scope are:

a. Drawn blood from vein or finger prick (comparable to blood glucose tests)

b. Cheek or other mucosal cells collected directly from source by swab

c. Tooth enamel

The TEI-REX program is not focused on detection of radionuclides absorbed, ingested, or injected into an organism.

The below categories of biomarkers are out of scope:

I. DNA damage to include single or double stranded breaks or associated repair signatures

II. Ratiometric expression profiling of DNA, RNA, and/or proteins

III. Biomarkers that require multiple collections from the same biological target (including baselines)

Radiation types out of scope for research activities under TEI-REX (TEI-REX capabilities can still be applicable):

I. Cosmic or heavy ions radiation

Final activities out of scope:

I. Manufacture of new equipment for analytical biomarker analysis.

II. Improvements on throughput, scalability, and multiplexing of existing biodosimetry approaches.

a. Establish approaches or biomarkers may be utilized in controls or confirmatory work only as a means to provide confidence in the model systems.

8 Program Test and Evaluation

TEI-REX will utilize independent T&E teams to assist in evaluating progress and success of TEI- REX approaches. Progress towards these milestones and equally weighted metrics are only one aspect of how program success will be monitored and assessed. The milestones and metrics are intended to focus and drive the TEI-REX program, while permitting flexibility, creativity, and innovation in the proposed solutions to meet the TEI-REX program goals. Proposals with a plan to surpass the listed milestone(s) and metrics are desirable and offerors will need to provide clear justification if their proposed approach will not be able to meet the listed milestones or metrics while still achieving the objectives of the program.

In addition to describing how proposed approaches address government-specific metrics, offerors should provide a detailed description of additional metrics or milestones relevant to their unique technical approaches. IARPA withholds the right to modify, remove, or add new milestones or metrics as the program progresses to ensure the research activities can be appropriately and effectively monitored and evaluated. Expected final Phase 1 metrics (Table 3) will be provided by the Government during the Phase 1 Kickoff Meeting. Any additional changes to milestones or metrics after program kickoff may be provided by the Government following discussions with program stakeholders.

The test and evaluation process includes three primary activities: 1) evaluation of self-reported scoring, against program metrics, by performer teams using their own samples throughout each Phase with progress reported through monthly deliverables; 2) evaluation of performer protocols, identified biomarker panels, and models by the T&E teams to substantiate self-reported scoring;

and 3) evaluation of performer biodosimetry and biomarker data outputs by T&E teams when evaluating blinded samples provided to performers by T&E. Dose exposure, timeline, and dose-rate associated with each blinded T&E sample will not be provided (refer to Table 3). Offerors should ensure they account for all described T&E events and processes and detail in their proposal any limitations or risks associated with their approach. All offerors should be aware that awarded proposals will be evaluated based upon how well their models and predictions meet program metrics when the ground truth of the blinded samples is released.

T&E is limited in the number of models and samples it can generate and offerors are encouraged to consider this when proposing appropriate model systems and sample types in their research pursuit of robust biomarkers. Proposed approaches which demonstrate very strong technical likelihood to meet the objective of TEI-REX while also falling outside of the model systems described will be considered but offerors must propose viable approaches for third-party test and evaluation approaches, enabling evaluation against program metrics, to account for the deviation.

TEI-REX T&E in Phases 1 and 2 will predominately leverage samples derived from in vivo mouse systems and full-thickness 3-D (mouse/human) constructs primarily derived from commercial sources. The Government, through the T&E Team(s), will provide up to 500 samples per T&E event. Events include Phase 1 Round 1, Phase 1 Round 2, and Phase 2 Round 3 (refer to Table 1 and Figure 2 for specific timing on T&E events); note that sample numbers are inclusive of experimental and biological replicates, controls, and standards to each team for each Focus Area.

T&E will include up to 25 unexposed negative controls per sample/model type per FA and up to 50 exposed samples with details of exposure prior to each T&E evaluation event (Table 1 and Figure 2) for performer self-assessment and developmental guidance. Many of these samples will likely be either skin, skin-like, or hair samples. If an offeror’s research focuses on additional sample types of interest that can be easily collected from either of these model types, the T&E team will attempt to provide a sufficient number of supporting, blinded samples to enable testing and evaluation, but this is not assured. If these alternative samples cannot be provided by T&E, performers will still be expected to meet program metrics by evaluating the samples provided.

Offerors should plan to receive two (2) sets of predominately blinded T&E samples in Phase 1.

The first set of samples, Phase 1 Round 1, will be used to establish a baseline capability of each performer’s approach against program metrics, while Phase 1 Round 2 will demonstrate how the teams have progressed across the Phase and their capability to achieve program metrics. Both sets of blinded samples will include at least one (1) sample type from one (1) model type. Each Focus Area will have its own set of samples. Offerors will be given 60 days to analyze all provided samples and submit their raw data and computational outputs from their models following guidelines provided by the Government after receipt of T&E samples. In Phase 1, performers are expected to determine, at least, the exposure dose and exposure timeline, to include error.

Biodosimetry model outputs must include error and interpretable association back to the biomarkers which informed towards the biodosimetry predictions. Biomarker detection metrics will align with true positive rate (TRP), false positive rate (FPR), and precision as statistical measurements as well as meeting sample mass/volume and extensibility requirements. Offerors are also expected to identify appropriate industry accepted quality score metrics or propose reasonable scoring approaches, based upon their analytical approach, which can be used to demonstrate successful biomarker detection from the matrix.

Offerors should plan that blinded Phase 2 T&E samples will include additional model or sample types, determined by program progress and performer successes in Phase 1, while still likely leveraging mouse and full-thickness 3-D constructs. Offerors will receive one (1) batch of predominately blinded samples, up to 500 samples, inclusive of blinded experimental and biological replicates, unblinded controls, and unblinded standards for each Focus Area. These samples will include at least two (2) sample types across two (2) model systems. Offerors will be given 60 days to analyze all samples provided and submit their raw data and outputs from their computational models. In Phase 2, metrics will expand from only scoring the prediction of dose exposure and exposure timeline to also evaluating the prediction of radiation exposure type and dose rate. Biodosimetry model outputs must include error and be interpretable, that is providing association to the biomarkers from which biodosimetry predictions were informed. Biomarker Detection metrics will align with true positive rate (TRP), false positive rate (FPR), and precision as statistical measurements as well as meeting sample mass/volume and extensibility requirements.

Offerors are also expected to identify appropriate industry accepted quality score metrics or propose reasonable scoring approaches, based upon their analytical approach, which can be used to demonstrate successful biomarker detection from the matrix.

Offerors should plan to receive multiple sets of a limited number of blinded samples across Phase 3, under 100 samples per Round. Phase 3 is aimed at researching and adapting the functional capabilities developed under Phases 1 and 2 to realistic samples. These samples will be directly informed through IARPA engagement with TEI-REX transition partners. The samples will often be limited in number and provide only empirical evidence of capability without strong statistical confidence, but performers will be expected to meet or exceed the sample metrics as required in Phase 2. These samples may include a wider range of model types, to potentially include samples from the Göttingen minipig, insects, human, and plants. These samples may include a range of confounders including: variations in age, gender, UV exposure, dose rate, and common chemicals typically used on hair or skin. During Phase 3, performers are expected to adapt and optimize their models appropriately throughout the three (3) rounds of the Phase while continuing to improve their platform and achieve Phase 3 metrics (Table 3).

Offerors should be prepared to use a reporting template, provided at the beginning of T&E activities in each Phase, to submit their results. This template will be developed by the T&E team(s) in coordination with performers. All supporting data derived during performer analysis after each T&E event will also be submitted for T&E review and program use.

Biodosimetry models developed by performers will be evaluated by T&E based upon selected metric outputs, interpretability of model outputs, and likelihood to meet overall TEI-REX objectives computed based on nonparametric statistical power studies to predict future model performance as a function of the number of available training samples.

Figure 2: Timeline for T&E sample activities across all three Phases

9 Government Provided Equipment and Samples To support the research and T&E processes, the T&E teams will provide multiple sets of government derived samples, to include unblinded negative control samples which are not irradiated and unblinded positive control standards that have been irradiated with a stepwise increase in dosage, within the relevant dose ranges, to support performer adaptation to T&E samples. Table 1, below, lists and describes these samples along with the expected metadata to be delivered with each sample set. The described metadata will be provided to support performer efforts as appropriate to the blinded samples being provided.

Table 1: Table of TEI-REX Provided Samples

Sample Name Phase/Activity Sample Description Sample Metadata Baseline (reference) samples

P1/M3 and P2/M20

Unexposed samples aligning with the T&E model systems and sample types to be tested in the respective phase and Focus Area. Up to 25 samples.

Model (age/gender/strain) Sample origin (hair/skin) Sample mass/volume

Unblinded training samples

P1/M7 and P2/M24

Irradiated samples aligning with T&E model systems and sample types provided prior to active T&E to enabled performers to test their biodosimetry approach in advance using 3rd party samples. Up to 50 samples.

Dose information (exposure/rate/timeline/ environment)

Model (age/gender/strain) Sample origin (hair/skin) Sample mass/volume

Blinded T&E samples

P1/M9, P1/M15, and P2/M28

Irradiated samples, primarily aligning with skin and/or hair, from T&E models. Blinded regarding all exposure information. Up to 500 samples.

Model (age/gender/strain) Sample origin

(hair/skin/etc.)

Sample mass/volume Spike-in concentrations Spike-in digestion profiles

T&E controls and standards

P1/M9, P1/M15, and P2/M28

Irradiated controls with known doses to create a positive control calibration curve.

Process control and platform performance standards spiked into or onto blinded samples to enable evaluation of sample prep by performers.

Dose information (exposure/rate/timeline/ environment)

Model (age/gender/strain) Sample origin (hair/skin) Sample mass/volume Spike-in concentrations Spike-in digestion profiles

Blinded realistic samples

P3/M31,35,39 Blinded irradiated samples derived from a variety of model systems aligning with exposure environments informed by transition partners. Up to 100 samples per round.

Sample origin (hair/skin/etc.)

Model system

10 Models, Software, and Application of Machine Learning/Artificial Intelligences Towards Biodosimetry

Offerors will be required to provide the algorithms, data, models, code, and/or software deliverables via a hosting environment established by the T&E team in a manner that conforms to industrial best practices, including containerized code to automate deployment. Offerors should describe how models will be developed, language(s) used, and expectations of command line or user interface development. Biodosimetry models will not be evaluated by hardware requirements but must be deployable on a cloud environment and packaged in a Docker or Singularity package with pre-identified dependencies. TEI-REX will not be providing an environment to train model systems and software.

The specifics of model and data delivery to the T&E environment will be provided at TEI-REX program Kickoff. Offerors should expect to utilize standardized reporting templates, likely in JSON or similar, with consistent and pre-identified terminology for labeling metadata tags.

Biodosimetry models are expected to incorporate both big data training from TEI-REX biomarkers and implicitly programmed elements focusing on known biodosimetry evaluation techniques, radiation physics, and biology. Predictions from the biodosimetry models must be interpretable, enabling review of the specific biomarkers leading towards the output results (Table 3).

Biodosimetry models must adhere with industry accepted coding standards and conventions and all models developed using machine learning/deep learning (ML/DL) must be retrainable by end-users.

T&E will leverage software deliverables to evaluate the functionality of the TEI-REX technologies, assess the extensibility of the technologies on different biomarker targets, sample types, and model types, and potentially ensemble multiple approaches and/or training databases developed under TEI-REX. This evaluation will be accomplished by T&E analyzing individual performer model outputs using raw biomarker data inputs generated across the program. Results of T&E verification should align with self-reported data from the performer teams and program metrics, any significant deviations will require further review of technologies by T&E with support from performer teams.

11 Program Waypoints, Milestones, and Metrics The Government will use the Program Goal (Section II, 1), Metrics (Table 3), and Expected Milestones and Deliverables (Table 4) as a partial means to determine whether satisfactory progress is being made to warrant continued funding for the program. The Government will use Milestones and Metrics to quantitatively track program progress, which for the purposes of this BAA are defined as:

Waypoint (Table 2): An intermediate performance target, tied to a specific time in the program execution, established by the performer but linked to a Milestone or Metric.

Milestone (Table 2; Figure 3): A specific Government-provided performance target, tied to a specific time in the program execution (e.g., establishment of irradiation testbed by Month 3). All milestones are required.

Metrics (Table 3): A quantitative or qualitative measure of program performance (e.g., prediction of exposure dose).

A Waypoint is a performer-specific performance target, tied to a specific time in the Program’s Execution (each performer will supply their own system- and approach-specific Waypoints).

Waypoints provide evidence that the technical and programmatic risks associated with the proposed approach are being addressed. Waypoints must be clear, well-defined, quantitative, and logically connected to offeror and/or Government decisions. Offerors must include Waypoints in their proposal and provide, as a deliverable, updates to the Program Waypoints at the start of each Phase. Performance against these waypoints will be reviewed throughout the Program to assess whether course corrections are needed to ensure Program success.

Program Milestones and Metrics define the scope and goals of the effort. The Government shall use the Program Milestones and Metrics, summarized in Tables 2 and 3, to evaluate the effectiveness of proposed solutions in achieving the stated program objectives, and in part to determine whether satisfactory progress is being made to warrant continued funding of the program. The offeror may also propose appropriate additional Milestones and Metrics to improve evaluation of progress. Additional Program Milestones should be proposed to provide evidence that the technical and programmatic risks associated with the proposed approach are being addressed. Any such Milestones and Metrics shall be clear and well-defined, with a logical connection to enabling offeror decisions and/or Government decisions.

The Metrics in Table 3 and overall constraints are intended to bound the scope of the effort, while affording maximum flexibility, creativity, and innovation in proposing solutions to the stated research problem. The TEI-REX program metrics are broken into two (2) types: 1) Metrics that are critical for ensuring the robustness of the biomarkers and their associated interpretability 2) Metrics that are critical for evaluation of effectiveness of biodosimetry models to characterize exposure incidents. These metrics may change as the Program progresses to ensure mission objectives are maintained while also continuing to drive innovation and growth within the Program. Most changes to metrics will occur following discussions between the TEI-REX IARPA team, ARO, TEI-REX transition partners, TEI-REX T&E team(s), and the performer(s) impacted by the change.

All Biomarker Detection metrics are associated with the analytical detection of biomarkers following radiation exposure. True positive rate (TPR) is the statistical measure used, as an aggregate across all possible biomarkers being targeted, with the goal of maximizing true detections. False positive rate (FPR) is the statistical measure used, as an aggregate across all possible biomarkers being targeted, with the goal of minimizing false detections. Precision is the statistical measure for measuring repeatability of biomarker detection, in aggregate. Extensibility is a binary call of yes or no, evaluated as achieving at least one (1) appropriate biomarker, used by the biodosimetry model to make predictions, in at least 50% of the new model systems. Finally, while multiplexing is not an objective of the final capabilities developed, we must consider how future applications of this research can be applied. As such, approaches need to develop capabilities which are timely when compared to other biodosimetry approaches, demonstrating continued improvement by decreasing analysis time throughout the program.

All Biodosimetry Model metrics are associated with the computational model developed by performers and its predictions, extensibility, and interpretability when evaluating TEI-REX and performer developed samples. All metrics should provide data associated with samples, based on the variable being tested, and all samples evaluated as a whole, as appropriate per model and sample type. Accuracy targets are defined as aggregated statistical measures for a given phases’ model predictions across dose, timeline, dose-rate, and ionizing radiation type. Median Absolute Error (MAE) is a statistical measure aimed to describe model prediction error. Precision is the statistical measure for measuring repeatability of model predictions across samples. Extensibility is a binary call of yes or no, defined by how well the biodosimetry model is able to predict the variable exposure factors when analyzing a new model type(s) with greater than 50% confidence across all samples. Interpretability is a retrospective correlative requiring the biodosimetry model to correlate predictions back to the biomarkers enabling the prediction to be made.

Waypoints and Milestones, developed by offerors and submitted with the proposal, shall include appropriate corresponding Metrics, and should be captured in a single table or timeline, similar to the structure found in Table 2, below.

Table 2: Template for Milestones and Waypoints, to be Completed by Offeror.

Phase Month(s) Event

(Milestone or Waypoint)

Description Comment Associated Deliverable or Metric

Table 3: Program Metrics Across All Phases and Focus Areas

Metric (FA1 and 2) Phase 1 Phase 2 Phase 3

Biomarker Detection

TPR 70% 80% 90%

FPR 30% 15% 10%

Precision 60% 70% 80%

Sample mass/volume analyzed < 50mg/50µL < 5mg/5µL < 5mg/5µL Extensibility: Biomarker(s) Detected in at Least X Sample Type(s) from X Model(s)

1 Sample / 1 Model

1 Sample / 2 Models

1 Sample / 3 Models

Analysis run-time 1 Sample / 24 hours

1 Sample / 10 hours

1 Sample / 3 hours

Biodosimetry Model

Prediction of Exposed Dose*

Accuracy 70% 80% 90%

MAE 30% 15% 10%

Precision 60% 70% 80%

Predicted Timing of Exposure (in days)*

Accuracy 70% 80% 90%

MAE 30% 15% 10%

Precision 60% 70% 80%

Prediction of Dose-Rate (mGray/min)*

Accuracy N/A 60% 70%

MAE N/A 30% 20%

Precision N/A 70% 80% Prediction of Ionizing Radiation Type (Particulate and/or Electromagnetic)*

Accuracy N/A 60% 80%

Precision N/A 70% 80%

Extensibility: Exposed Dose Predicted in at Least X Sample Type(s) from X Model(s)

1 Sample / 1 Model

1 Sample / 2 Models

1 Sample / 3 Models

Model Interpretability: Model identifies X percent of composite biomarkers informing towards predictions

70% 80% 90%

*must be evaluated at the grouped level of sample evaluation (target variable is held constant) as well across the entire collection of samples evaluated

12 Program Period of Performance, Timeline, and Deliverables

IARPA will use the timelines as described in Figure 3 and Table 4 to monitor, evaluate, and maintain overall Program progress and its 42-month Program Schedule. Offerors should plan for a 42-month effort for the Period of Performance over three (3) phases: Phase 1 of TEI-REX shall last 18 months; Phase 2 shall last 12 months, and Phase 3 shall last 12 months. Decisions for Phase 2 and 3 options will be based on successfully meeting program goals in the previous phases and funding availability. Refer to Figure 3 for a more complete capture of Program timeline and activities. Table 3 includes a schedule for the key deliverables the offerors shall provide. In addition to technical oversight of progress, technical reviews will assess programmatic progress against proposed work plans. Offerors may add additional deliverables as needed to the minimum set listed in Table 3.

Table 4: Expected Milestones and Deliverables Associated with TEI-REX Phase/Month Milestone/Deliverable

P1/M1, P2/M19, P3/M31

Program/Phase kickoff meetings, likely in the Washington DC Metropolitan area. Drafts of presentation material are due 5 workdays prior to kickoff. Final materials due 15 calendar days following meeting date.

P1/M1, P2/M19, P3/M31

Processed approvals, as appropriate for Phase 1 activities, to include IRB and IACUC review. These approvals shall be updated annually or as appropriate.

P1/M1 Updated Phase 1 research plan to include waypoints…