Probabilistic Assignment Research Articles

Equity-centered adaptive sampling in sub-sewershed wastewater surveillance using census data.

Sub-city, or sub-sewershed, wastewater monitoring for infectious diseases offers a data-driven strategy to inform local public health response and complements city-wide data from centralized wastewater treatment plants. Developing strategies for equitable representation of diverse populations in sub-city wastewater sampling frameworks is complicated by misalignment between demographic data and sampling zones. We address this challenge by: (1) developing a geospatial analysis tool that probabilistically assigns demographic data for subgroups aggregated by race and age from census blocks to sub-city sampling zones; (2) evaluating representativeness of subgroup populations for COVID-19 wastewater-based disease surveillance in Davis, California; and (3) demonstrating scenario planning that prioritizes vulnerable populations. We monitored SARS-CoV-2 in wastewater as a proxy for COVID-19 incidence in Davis (November 2021-September 2022). Daily city-wide sampling and thrice-weekly sub-city sampling from 16 maintenance holes covered nearly the entire city population. Sub-city wastewater data, aggregated as a population-weighted mean, correlated strongly with centralized treatment plant data (Spearman's correlation 0.909). Probabilistic assignment of demographic data can inform decisions when adapting sampling locations to prioritize vulnerable groups. We considered four scenarios that reduced the number of sampling zones from baseline by 25% and 50%, chosen randomly or to prioritize coverage of >65-year-old populations. Prioritizing representation increased coverage of >65-year-olds from 51.1% to 67.2% when removing half the zones, while increasing coverage of Black or African American populations from 67.5% to 76.7%. Downscaling had little effect on correlations between sub-city and centralized data (Spearman's correlations ranged from 0.875 to 0.917), with strongest correlations observed when prioritizing coverage of >65-year-old populations.

Genetic homogeneity and weak signatures of local adaptation in the marine mussel Mytilus chilensis

The natural populations of the marine mussel Mytilus chilensis and the associated aquaculture industry forms a sensitive social-ecological system that relies on the released propagules for cultivation in the highly heterogeneous environment (temperature, productivity, and salinity) of northern Patagonia (42–44 °S). We assessed spatial genetic structure, signals of local adaptation, and population assignment of M. chilensis analyzing 5963 SNPs from 125 individuals across six natural populations sampled over two consecutive years along the southeast Pacific coast (39° 25′ to 43° 07′ S, ~ 430 km). Neutral and putatively adaptive loci revealed high genetic diversity and low genetic differentiation among populations. Of the whole dataset, less than 1% (50) of loci were identified as putatively adaptive through multiple approaches, with only 0.1% detected in by all of them, and only two loci of them were correlated with environmental variables. No evidence of Isolation by Environment (IBE) was found, albeit a slight differentiation in the southern sampling location (Yaldad). These results suggest that the genetic structure observed is primarily shaped by neutral processes with weak signals of local adaptation. Gene-flow appears to be the main evolutionary force influencing the species’ population genetic structure. Because of the importance for the industry, the probability of correct assignment of individuals to their population of origin using allelic frequencies was evaluated. Analyses exhibited relatively low probabilities (< 50% for four out of six sites) of accurately assigning individuals to their geographic origin, with a limited success of SNP markers the for such purposes. Likely, species' high dispersal capacity, seed translocation, and the spill-over effect of mussel aquaculture prevents population genetic differentiation through high effective gene flow, hindering local genetic adaptation.

A scalable approach to topic modelling in single-cell data by approximate pseudobulk projection.

Probabilistic topic modelling has become essential in many types of single-cell data analysis. Based on probabilistic topic assignments in each cell, we identify the latent representation of cellular states. A dictionary matrix, consisting of topic-specific gene frequency vectors, provides interpretable bases to be compared with known cell type-specific marker genes and other pathway annotations. However, fitting a topic model on a large number of cells would require heavy computational resources-specialized computing units, computing time and memory. Here, we present a scalable approximation method customized for single-cell RNA-seq data analysis, termed ASAP, short for Annotating a Single-cell data matrix by Approximate Pseudobulk estimation. Our approach is more accurate than existing methods but requires orders of magnitude less computing time, leaving much lower memory consumption. We also show that our approach is widely applicable for atlas-scale data analysis; our method seamlessly integrates single-cell and bulk data in joint analysis, not requiring additional preprocessing or feature selection steps.

Systematic Effects in Galaxy-Galaxy Lensing with DESI

The Dark Energy Spectroscopic Instrument (DESI) survey will measure spectroscopic redshifts for millions of galaxies across roughly 14,000deg2 of the sky. Cross-correlating targets in the DESI survey with complementary imaging surveys allows us to measure and analyze shear distortions caused by gravitational lensing in unprecedented detail. In this work, we analyze a series of mock catalogs with ray-traced gravitational lensing and increasing sophistication to estimate systematic effects on galaxy-galaxy lensing estimators such as the tangential shear γt and the excess surface density ΔΣ. We employ mock catalogs tailored to the specific imaging surveys overlapping with the DESI survey: the Dark Energy Survey (DES), the Hyper Suprime-Cam (HSC) survey, and the Kilo-Degree Survey (KiDS). Among others, we find that fiber incompleteness can have significant effects on galaxy-galaxy lensing estimators but can be corrected effectively by up-weighting DESI targets with fibers by the inverse of the fiber assignment probability. Similarly, we show that intrinsic alignment and lens magnification are expected to be statistically significant given the precision forecasted for the DESI year-1 data set. Our study informs several analysis choices for upcoming cross-correlation studies of DESI with DES, HSC, and KiDS.

Probabilistic Assignment With Decoupled IoU Prediction for Visual Tracking

Probabilistic Assignment With Decoupled IoU Prediction for Visual Tracking

Financial incentives to motivate treatment for hepatitis C with direct acting antivirals among Australian adults (The Methodical evaluation and Optimisation of Targeted IncentiVes for Accessing Treatment of Early-stage hepatitis C: MOTIVATE-C): protocol for a dose-response randomised controlled study

BackgroundUntreated hepatitis C virus (HCV) infection can result in cirrhosis and hepatocellular cancer. Direct-acting antiviral (DAA) therapies are highly effective and have few side effects compared to older interferon-based therapy. Despite the Australian government providing subsidised and unrestricted access to DAA therapy for chronic HCV infection, uptake has not been sufficient to meet the global target of eliminating HCV as a public health threat by 2030. This study will offer people with HCV financial incentives of varying values in order to evaluate its effect on initiation of DAA therapy in primary care.MethodsAustralian adults (18 years or older) who self-report as having current untreated HCV infection can register to participate via an automated SMS-based system. Following self-screening for eligibility, registrants are offered a financial incentive of randomised value (AUD 0 to 1000) to initiate DAA therapy. Study treatment navigators contact registrants who have consented to be contacted, to complete eligibility assessment, outline the study procedures (including the requirement for participants to consult a primary care provider), obtain consent, and finalise enrolment. Enrolled participants receive their offered incentive on provision of evidence of DAA therapy initiation within 12 weeks of registration (primary endpoint). Balanced randomisation is used across the incentive range until the first analysis, after which response-adaptive randomisation will be used to update the assignment probabilities. For the primary analysis, a Bayesian 4-parameter EMAX model will be used to estimate the dose–response curve and contrast treatment initiation at each incentive value against the control arm (AUD 0). Specified secondary statistical and economic analyses will evaluate the effect of incentives on adherence to DAA therapy, virological response, and cost-effectiveness.DiscussionThis project seeks to gain an understanding of the dose–response relationship between incentive value and DAA treatment initiation, while maximising the number of people treated for HCV within fixed budget and time constraints. In doing so, we hope to offer policy-relevant recommendation(s) for the use of financial incentives as a pragmatic, efficient, and cost-effective approach to achieving elimination of HCV from Australia.Trial registrationANZCTR (anzctr.org.au), Identifier ACTRN12623000024640, Registered 11 January 2023 (https://anzctr.org.au/Trial/Registration/TrialReview.aspx?id=384923&isReview=true).

Distribution-free Inferential Models: Achieving finite-sample valid probabilistic inference, with emphasis on quantile regression

This paper presents a novel distribution-free Inferential Model (IM) construction that provides valid probabilistic inference across a broad spectrum of distribution-free problems, even in finite sample settings. More specifically, the proposed IM has the capability to assign (imprecise) probabilities to assertions of interest about any feature of the unknown quantities under examination, and these probabilities are well-calibrated in a frequentist sense. It is also shown that finite-sample confidence regions can be derived from the IM for any such features. Particular emphasis is placed on quantile regression, a domain where uncertainty quantification often takes the form of set estimates for the regression coefficients in applications. Within this context, the IM facilitates the acquisition of these set estimates, ensuring they are finite-sample confidence regions. It also enables the provision of finite-sample valid probabilistic assignments for any assertions of interest about the regression coefficients. As a result, regardless of the type of uncertainty quantification desired, the proposed framework offers an appealing solution to quantile regression.

PROTAX-GPU: a scalable probabilistic taxonomic classification system for DNA barcodes.

DNA-based identification is vital for classifying biological specimens, yet methods to quantify the uncertainty of sequence-based taxonomic assignments are scarce. Challenges arise from noisy reference databases, including mislabelled entries and missing taxa. PROTAX addresses these issues with a probabilistic approach to taxonomic classification, advancing on methods that rely solely on sequence similarity. It provides calibrated probabilistic assignments to a partially populated taxonomic hierarchy, accounting for taxa that lack references and incorrect taxonomic annotation. While effective on smaller scales, global application of PROTAX necessitates substantially larger reference libraries, a goal previously hindered by computational barriers. We introduce PROTAX-GPU, a scalable algorithm capable of leveraging the global Barcode of Life Data System (>14 million specimens) as a reference database. Using graphics processing units (GPU) to accelerate similarity and nearest-neighbour operations and the JAX library for Python integration, we achieve over a 1000 × speedup compared with the central processing unit (CPU)-based implementation without compromising PROTAX's key benefits. PROTAX-GPU marks a significant stride towards real-time DNA barcoding, enabling quicker and more efficient species identification in environmental assessments. This capability opens up new avenues for real-time monitoring and analysis of biodiversity, advancing our ability to understand and respond to ecological dynamics. This article is part of the theme issue 'Towards a toolkit for global insect biodiversity monitoring'.

Ex ante transparency and corruption by networks

AbstractThis paper explains the structure of corruption networks as response to ex ante transparency, defined as visibility of authorities whose cooperation clients may need, in due course, to execute corrupt transactions. It also characterizes the optimal transparency policy given the network response, as a function of connection costs, sanctions, the corruption surplus and the detection probability (hence the anti‐corruption budget and ex post transparency). Corruption chains may emerge in equilibrium if authority is expected to be shared by multiple offices, where the office with higher solo assignment probability becomes the intermediary. Otherwise, clients penetrate the bureaucracy by inducing the star network, or contend with single connection. I show that the optimal policy always assigns one office, sometimes alone, sometimes jointly with others. It is often possible to deter corruption networks through an ex ante transparent policy that parcels out authority to multiple offices with probability 1—a common feature of many US bureaucracies. Decomposing transparency into its components reveals nuances in the transparency–corruption relationship, suggesting that ex ante transparency is instrumental except in environments in which anti‐corruption enforcement is extremely effective or extremely ineffective.

Population assignment of migratory Westslope Cutthroat Trout (WCT) in the Clark Fork–Pend Oreille River basin

AbstractObjectiveThe Clark Fork–Pend Oreille River basin of northeastern Washington and the Idaho Panhandle historically supported a robust metapopulation of the Westslope Cutthroat Trout (WCT) Oncorhynchus lewisi, a western native salmonid of high cultural and economic value. The construction of impassible hydroelectric dams and smaller instream barriers has prevented the return of migratory WCT to spawning tributaries, leading to the fragmentation of this metapopulation over the past 100 years. One such impassible barrier is Albeni Falls Dam (AFD) near Newport, Washington, which was completed without fish passage capabilities in 1955. We sought to examine large‐scale genetic patterns in the study area and determine the most likely spawning tributary of origin for migratory WCT captured below AFD.MethodsWe created a genetic baseline representative of populations within the Clark Fork–Pend Oreille River basin from upstream and downstream of the dam using 191 biallelic single‐nucleotide polymorphism genetic markers. Our data set included 124 collections, which allowed for an examination of population structure and hatchery influence across the study area and provided a robust tool for population assignment. Population assignment tests were conducted using the program RUBIAS.ResultPopulation assignment tests were successful for all pure WCT of unknown origin despite potential influence from hatchery lineages across the study area. Of 83 migratory WCT captured below AFD, approximately 80% were assigned to tributaries upstream of AFD with a posterior assignment probability of at least 90%. Only one fish was assigned to a tributary downstream of AFD.ConclusionOur results indicate that AFD disrupts the natural metapopulation dynamics of WCT populations in the basin. Passage for WCT at this barrier would reestablish metapopulation connectivity within the basin by allowing migratory individuals to make genetic contributions to populations upstream of the dam.

145 Encoding Brain Dynamics: Neuropsychological Test Scores as Predictors of iEEG Spectral Features in Epilepsy

INTRODUCTION: Intracranial EEG is a valuable tool for investigating the neural mechanisms underlying medically refractory epilepsy (MRE). Despite the advances made in understanding the condition, there is a lack of research exploring the relationship between iEEG data and neuropsychological test performance in MRE patients. METHODS: We performed a retrospective analysis of resting-state iEEG data from 70 MRE patients, extracting spectral features from 4.4 million 10-second time windows across 5201 electrodes. Using MATLAB and FieldTrip toolbox, we processed the data and used two brain atlases for probabilistic electrode assignment. Neuropsychological test scores were corrected for demographic factors. Relationships between preoperative neuropsychological test scores and spectral features were explored using nested hierarchical generalized linear mixed effects (GLME) models, focusing on the hippocampus. RESULTS: We found significant associations between neuropsychological tests and hippocampal power spectral densities (PSDs). Rey Auditory verbal learning test (RAVLT) recognition correct and short-term retention percentages were significantly associated with relative power (RP) in the beta-range (p &lt; 0.05, FDR-corrected). The Boston naming test and Trail Making Test (TMT)-A finishing time showed significant associations with high-beta/low-gamma PSD (p &lt; 0.05, FDR-corrected). TMT-B errors are positively correlated with delta/theta RP (p &lt; 0.05, FDR-corrected). CONCLUSIONS: Our study provides novel insights into the relationship between iEEG spectral features and neuropsychological test performance in MRE patients. The findings suggest that the heterogeneity in spectral features may explain relevant cognitive performance, highlighting the biological basis and circuits of neuropsychological testing.

Robustness of response-adaptive randomization.

Doubly adaptive biased coin design (DBCD), a response-adaptive randomization scheme, aims to skew subject assignment probabilities based on accrued responses for ethical considerations. Recent years have seen substantial advances in understanding DBCD's theoretical properties, assuming correct model specification for the responses. However, concerns have been raised about the impact of model misspecification on its design and analysis. In this paper, we assess the robustness to both design model misspecification and analysis model misspecification under DBCD. On one hand, we confirm that the consistency and asymptotic normality of the allocation proportions can be preserved, even when the responses follow a distribution other than the one imposed by the design model during the implementation of DBCD. On the other hand, we extensively investigate three commonly used linear regression models for estimating and inferring the treatment effect, namely difference-in-means, analysis of covariance (ANCOVA) I, and ANCOVA II. By allowing these regression models to be arbitrarily misspecified, thereby not reflecting the true data generating process, we derive the consistency and asymptotic normality of the treatment effect estimators evaluated from the three models. The asymptotic properties show that the ANCOVA II model, which takes covariate-by-treatment interaction terms into account, yields the most efficient estimator. These results can provide theoretical support for using DBCD in scenarios involving model misspecification, thereby promoting the widespread application of this randomization procedure.

Abstract 7415: Assigning mutational signatures to individual samples and individual somatic mutations with SigProfilerAssignment

Abstract Analysis of mutational signatures is a powerful approach for understanding the mutagenic processes that have shaped the evolution of a cancer genome. To evaluate the mutational signatures operative in a cancer genome, one first needs to quantify their activities by estimating the number of mutations imprinted by each signature, a process known as refitting of mutational signatures. Although multiple tools for refitting known signatures were developed in the past decade, most provide support almost exclusively for single base substitution signatures and lack an online interface. Further, these tools have never been compared, and no existing tool supports probabilistically assigning signatures to a single somatic mutation. Here, we present SigProfilerAssignment, a desktop and online computational framework for assigning all types of mutational signatures to individual samples. SigProfilerAssignment is the first tool that allows both analysis of copy-number signatures and probabilistic assignment of signatures to individual somatic mutations. As its computational engine, the tool uses a custom implementation of the forward stagewise algorithm for sparse regression and nonnegative least squares for numerical optimization. The analysis of 2,700 synthetic cancer genomes with and without noise demonstrates that SigProfilerAssignment outperforms four commonly used approaches for assigning mutational signatures, with consistent top performance across cancer types and most signatures. SigProfilerAssignment stands out as the most precise and sensitive tool while maintaining high computational performance and bringing novel capabilities. SigProfilerAssignment is distributed as an open-source Python/R package available at https://github.com/AlexandrovLab/SigProfilerAssignment, with an online interface available as part of the COSMIC Mutational Signatures website at https://cancer.sanger.ac.uk/signatures/assignment/. Citation Format: Raviteja Vangara, Marcos Diaz-Gay, Mark D. Barnes, Xi Wang, S M Ashiqul Islam, Ian Vermes, Stephen Duke, Nithish Bharadhwaj Narasimman, Ting Yang, Zichen Jiang, Sarah Moody, Sergey Senkin, Paul Brennan, Michael R. Stratton, Ludmil B. Alexandrov. Assigning mutational signatures to individual samples and individual somatic mutations with SigProfilerAssignment [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7415.

Explainable AI to identify radiographic features of pulmonary edema

Abstract Background Pulmonary edema is a leading cause for requiring hospitalization in patients with congestive heart failure. Assessing the severity of this condition with radiological imaging becomes paramount in determining the optimal course of patient care. Purpose This study aimed to develop a deep learning methodology for the identification of radiographic features associated with pulmonary edema. Materials and Methods This retrospective study used a dataset from the Medical Information Mart for Intensive Care database comprising 1000 chest radiograph images from 741 patients with suspected pulmonary edema. The images were annotated by an experienced radiologist, who labeled radiographic manifestations of cephalization, Kerley lines, pleural effusion, bat wings, and infiltrate features of edema. The proposed methodology involves 2 consecutive stages: lung segmentation and edema feature localization. The segmentation stage is implemented using an ensemble of 3 networks. In the subsequent localization stage, we evaluated 8 object detection networks, assessing their performance with average precision (AP) and mean AP. Results Effusion, infiltrate, and bat wing features were best detected by the Side-Aware Boundary Localization (SABL) network with corresponding APs of 0.599, 0.395, and 0.926, respectively. Furthermore, SABL achieved the highest overall mean AP of 0.568. The Cascade Region Proposal Network network attained the highest AP of 0.417 for Kerley lines and the Probabilistic Anchor Assignment network achieved the highest AP of 0.533 for cephalization. Conclusion The proposed methodology, with the application of SABL, Cascade Region Proposal Network, and Probabilistic Anchor Assignment detection networks, is accurate and efficient in localizing and identifying pulmonary edema features and is therefore a promising diagnostic candidate for interpretable severity assessment of pulmonary edema.

Battling the coronavirus 'infodemic' among social media users in Kenya and Nigeria.

How can we induce social media users to be discerning when sharing information during a pandemic? An experiment on Facebook Messenger with users from Kenya (n = 7,498) and Nigeria (n = 7,794) tested interventions designed to decrease intentions to share COVID-19 misinformation without decreasing intentions to share factual posts. The initial stage of the study incorporated: (1) a factorial design with 40 intervention combinations; and (2) a contextual adaptive design, increasing the probability of assignment to treatments that worked better for previous subjects with similar characteristics. The second stage evaluated the best-performing treatments and a targeted treatment assignment policy estimated from the data. We precisely estimate null effects from warning flags and related article suggestions, tactics used by social media platforms. However, nudges to consider the accuracy of information reduced misinformation sharing relative to control by 4.9% (estimate = -2.3 percentage points, 95% CI = [-4.2, -0.35]). Such low-cost scalable interventions may improve the quality of information circulating online.

Improving Estimation Efficiency via Regression-Adjustment in Covariate-Adaptive Randomizations with Imperfect Compliance

Abstract We investigate how to improve efficiency using regression adjustments with covariates in covariate-adaptive randomizations (CARs) with imperfect subject compliance. Our regression-adjusted estimators, which are based on the doubly robust moment for local average treatment effects, are consistent and asymptotically normal even with heterogeneous probabilities of assignment and misspecified regression adjustments. We propose an optimal but potentially misspecified linear adjustment and its further improvement via a nonlinear adjustment, both of which lead to more efficient estimators than the one without adjustments. We also provide conditions for nonparametric and regularized adjustments to achieve the semiparametric efficiency bound under CARs.

A Self-Organization Formation Configuration Based Assignment Probability and Collision Detection

A Self-Organization Formation Configuration Based Assignment Probability and Collision Detection

Comparison of Pocock and Simon’s covariate-adaptive randomization procedures in clinical trials

When multiple influential covariates need to be balanced during a clinical trial, stratified blocked randomization and covariate-adaptive randomization procedures are frequently used in trials to prevent bias and enhance the validity of data analysis results. The latter approach is increasingly used in practice for a study with multiple covariates and limited sample sizes. Among a group of these approaches, the covariate-adaptive procedures proposed by Pocock and Simon are straightforward to be utilized in practice. We aim to investigate the optimal design parameters for the patient treatment assignment probability of their developed three methods. In addition, we seek to answer the question related to the randomization performance when additional covariates are added to the existing randomization procedure. We conducted extensive simulation studies to address these practically important questions.

DOP17 Tofacitinib versus Vedolizumab Among Bio-naïve Patients With Ulcerative Colitis: A Real-World Propensity-Weighted Comparison

Abstract Background Over the last decade, treatment options for moderate-to-severe ulcerative colitis (UC) have expanded. However, comparative studies between these agents are limited, especially among biologic naïve patients. There are sparse data comparing biologics with small molecules for IBD. We aimed to compare the efficacy, safety and persistence of tofacitinib and vedolizumab as the first advanced treatment for patients with UC. Methods Patients who received tofacitinib or vedolizumab as first advanced therapy for UC in NHS Lothian were included. We capped the upper age limit at 65 years to take into account regulatory guidelines for the use of JAKi as first line therapies across IMIDs. To allow treatment effect to be assessed outside of a randomized trial, we used inverse probability of treatment weighting (IPTW). This approach takes the probability of treatment assignment into account without potentially drastically reducing the analysable cohort size. The probability of treatment assignment was calculated via logistic regression using age, gender, IBD duration, Montreal extent, CRP, concomitant corticosteroids and partial Mayo score at drug commencement. Missing data for these variables were imputed using multivariate imputation by chained equations. Confounder-adjusted survival curves were created using Kaplan-Meier estimates weighted via IPTW. The Pepe and Fleming test was used to compare survival curves from drug commencement to day 1000 of treatment. Results We included 158 patients of whom 81 (51.2%) received vedolizumab and 77 (48.7%) tofacitinib. Median follow-up for patients on vedolizumab was 3.1 (1.6-4.8) years and for tofacitinib 1.5 (0.34-2.3) years. Baseline demographics were comparable except for disease extent (Table 1). At drug commencement there were no differences in steroid prescription (60.5% vedolizumab versus 57.1% tofacitinib), partial Mayo score, CRP or faecal calprotectin. At week 12, steroid free clinical remission was more frequent in the vedolizumab group (69% vs 51.4%, p=0.030) Figure 1B. Vedolizumab persistence was superior to tofacitinib (p=0.005) Figure 1A. Primary non-response and secondary loss of response were 9.9% and 17.3% for vedolizumab and 23.4% and 13% for tofacitinib respectively. There were no differences in the frequency of adverse events (11 [13.6%] vedolizumab vs 19 [24.7%] tofacitinib, p=0.629). However, adverse events resulting in temporary discontinuation of the drug occurred in 1 (1.2%) vedolizumab versus 11 (14.3%) tofacitinib, p=0.013. Conclusion We found that the persistence and tolerability of vedolizumab was superior to tofacitinib in bionaive UC, although the rates of clinical and biomarker remission were comparable. These data may help inform positioning of advanced therapy in UC.

Probabilistic Hesitant Fuzzy Evidence Theory and Its Application in Capability Evaluation of a Satellite Communication System.

Evaluating the capabilities of a satellite communication system (SCS) is challenging due to its complexity and ambiguity. It is difficult to accurately analyze uncertain situations, making it difficult for experts to determine appropriate evaluation values. To address this problem, this paper proposes an innovative approach by extending the Dempster-Shafer evidence theory (DST) to the probabilistic hesitant fuzzy evidence theory (PHFET). The proposed approach introduces the concept of probabilistic hesitant fuzzy basic probability assignment (PHFBPA) to measure the degree of support for propositions, along with a combination rule and decision approach. Two methods are developed to generate PHFBPA based on multi-classifier and distance techniques, respectively. In order to improve the consistency of evidence, discounting factors are proposed using an entropy measure and the Jousselme distance of PHFBPA. In addition, a model for evaluating the degree of satisfaction of SCS capability requirements based on PHFET is presented. Experimental classification and evaluation of SCS capability requirements are performed to demonstrate the effectiveness and stability of the PHFET method. By employing the DST framework and probabilistic hesitant fuzzy sets, PHFET provides a compelling solution for handling ambiguous data in multi-source information fusion, thereby improving the evaluation of SCS capabilities.