Inference Tools Research Articles

A machine learning approach for estimating Eastern Asian origins from massive screening of Y chromosomal short tandem repeats polymorphisms.

Inferring the ancestral origin of DNA evidence recovered from crime scenes is crucial in forensic investigations, especially in the absence of a direct suspect match. Ancestry informative markers (AIMs) have been widely researched and commercially developed into panels targeting multiple continental regions. However, existing forensic ancestry inference panels typically group East Asian individuals into a homogenous category without further differentiation. In this study, we screened Y chromosomal short tandem repeat (Y-STR) haplotypes from 10,154 Asian individuals to explore their genetic structure and generate an ancestry inference tool through a machine learning (ML) approach. Our research identified distinct genetic separations between East Asians and their neighboring Southwest Asians, with tendencies of northern and southern differentiation observed within East Asian populations. All machine learning models developed in this study demonstrated high accuracy, with the Asian classification model achieving an optimal performance of 82.92% and the East Asian classification model reaching 84.98% accuracy. This work not only deepens the understanding of genetic substructures within Asian populations but also showcases the potential of ML in forensic ancestry inference using extensive Y-STR data. By employing computational methods to analyze intricate genetic datasets, we can enhance the resolution of ancestry in forensic contexts involving Asian populations.

What Can Economics Contribute to the Study of Terrorism? – The Power and Limitation of Economics

ABSTRACT We assess the potential of economics to contribute to the understanding of terrorism. We argue that the economic approach portraying terrorists and law enforcement agents as rational actors yoked to other disciplines is a source of precise falsifiable causal hypotheses. Theoretical economics rooted in rational choice provides precisely falsifiable hypotheses. This falsifiability makes it easier to focus on identifying causal relationships, which is essential for sound policy advice. Economists have developed a multitude of rigorous tools for causal inference ranging from instrumental variables to machine learning, which are essential in typical situations of terrorism that are beset by endogeneity problems. However, neighboring disciplines are needed to provide a rich contextual analysis that allows the assessment of the external validity of the results rigorously derived by economic analyses. Ultimately, rigorous mathematical analysis is valid only if it can capture reality wholly and simply.

Inferring Distant Relationships From Dense SNP Data Utilizing Two Genealogy Algorithms.

A highly esteemed method known as investigative genetic genealogy (IGG) has been developed to identify DNA samples from forensic crime scenes and human remains of disaster victims. With the advent of next-generation sequencing, it is now feasible to access information on millions of SNPs typed in a single sequencing run that fulfill the requirements for kinship inference. However, challenges such as the poor quality of forensic samples, the high cost associated with sequencing technology, and privacy concerns regarding large-scale genetic databases remain unresolved in this field. In the present study, we validated the identification of relationships up to the seventh degree using two genealogy algorithms (IBIS and KING) under various parameter settings. This was accomplished through whole genome sequencing data derived from two southern Chinese Han pedigrees during an initial phase, while also exploring workflows adapted for low-quality samples. To achieve this objective, low-coverage whole genome sequencing data were downsampled from high-coverage original datasets; additionally, mimic SNP array data-containing less information but offering greater accessibility-were prepared as reference samples. Through a series of experimental analyses, we not only validate the applicability of selected processing procedures and inference tools for low-coverage samples but also proposed that a meticulously crafted site filtering strategy can significantly improve the accuracy of kinship identification. This acknowledges the necessity for further systematic evidence in future research endeavors.

SCARAP: scalable cross-species comparative genomics of prokaryotes.

Much of prokaryotic comparative genomics currently relies on two critical computational tasks: pangenome inference and core genome inference. Pangenome inference involves clustering genes from a set of genomes into gene families, enabling genome-wide association studies and evolutionary history analysis. The core genome represents gene families present in nearly all genomes and is required to infer a high-quality phylogeny. For species-level datasets, fast pangenome inference tools have been developed. However, tools applicable to more diverse datasets are currently slow and scale poorly. Here, we introduce SCARAP, a program containing three modules for comparative genomics analyses: a fast and scalable pangenome inference module, a direct core genome inference module and a module for subsampling representative genomes. When benchmarked against existing tools, the SCARAP pan module proved up to an order of magnitude faster with comparable accuracy. The core module was validated by comparing its result against a core genome extracted from a full pangenome. The sample module demonstrated the rapid sampling of genomes with decreasing novelty. Applied to a dataset of over 31,000 Lactobacillales genomes, SCARAP showcased its ability to derive a representative pangenome. Finally, we applied the novel concept of gene fixation frequency to this pangenome, showing that Lactobacillales genes that are prevalent but rarely fixate in species often encode bacteriophage functions. The SCARAP toolkit is publicly available at https://github.com/swittouck/scarap. Supplementary data are available at Bioinformatics online.

Exploring spectropolarimetric inversions using neural fields. Solar chromospheric magnetic field under the weak-field approximation

Full-Stokes polarimetric datasets, originating from slit-spectrograph or narrow-band filtergrams, are routinely acquired nowadays. The data rate is increasing with the advent of bi-dimensional spectropolarimeters and observing techniques that allow long-time sequences of high-quality observations. There is a clear need to go beyond the traditional pixel-by-pixel strategy in spectropolarimetric inversions by exploiting the spatiotemporal coherence of the inferred physical quantities that contain valuable information about the conditions of the solar atmosphere. We explore the potential of neural networks as a continuous representation of the physical quantities over time and space (also known as neural fields), for spectropolarimetric inversions. We have implemented and tested a neural field to perform one of the simplest forms of spectropolarimetric inversions, the inference of the magnetic field vector under the weak-field approximation (WFA). By using a neural field to describe the magnetic field vector, we regularized the solution in the spatial and temporal domain by assuming that the physical quantities are continuous functions of the coordinates. This technique can be trivially generalized to account for more complex inversion methods. We have tested the performance of the neural field to describe the magnetic field of a realistic 3D magnetohydrodynamic (MHD) simulation. We have also tested the neural field as a magnetic field inference tool (approach also known as physics-informed neural networks) using the WFA as our radiative transfer model. We investigated the results in synthetic and real observations of the Ca ii line. We also explored the impact of other explicit regularizations, such as using the information of an extrapolated magnetic field, or the orientation of the chromospheric fibrils. Compared to traditional pixel-by-pixel inversions, the neural field approach improves the fidelity of the reconstruction of the magnetic field vector, especially the transverse component. This implicit regularization is a way of increasing the effective signal to noise of the observations. Although it is slower than the pixel-wise WFA estimation, this approach shows a promising potential for depth-stratified inversions, by reducing the number of free parameters and inducing spatiotemporal constraints in the solution.

Bayesian Analysis of the Generalized Additive Proportional Hazards Model: Asymptotic Studies

In this paper, we study Bayesian asymptotic properties of the proportional hazards model where the link function is modeled by the generalized additive model. As the standard generalized additive model is, the generalized additive proportional hazards model is a useful tool in finding the nonlinearity of covariate effects to survival times. We develop a data-dependent sieve prior for the generalized additive link function and use the Bayesian bootstrap prior for the baseline hazard function. We prove that the posterior contraction rate of the generalized additive link function is minimax optimal up to a logn term when the prior is carefully selected. By analyzing simulated as well as real data, we verify our theoretical results and compare with exisiting algorithms for the generalized additive proportional hazards model to illustrate that the proposed Bayesian model is a useful inference tool.

Spread.gl: visualising pathogen dispersal in a high-performance browser application.

Bayesian phylogeographic analyses are pivotal in reconstructing the spatio-temporal dispersal histories of pathogens. However, interpreting the complex outcomes of phylogeographic reconstructions requires sophisticated visualisation tools. To meet this challenge, we developed spread.gl, an open-source, feature-rich browser application offering a smooth and intuitive visualisation tool for both discrete and continuous phylogeographic inferences, including the animation of pathogen geographic dispersal through time. Spread.gl can render and combine the visualisation of multiple layers that contain information extracted from the input phylogeny and diverse environmental data layers, enabling researchers to explore which environmental factors may have impacted pathogen dispersal patterns before conducting formal testing. We showcase the visualisation features of spread.gl with representative examples, including the smooth animation of a phylogeographic reconstruction based on more than 17,000 SARS-CoV-2 genomic sequences. Source code, installation instructions, example input data, and outputs of spread.gl are accessible at https://github.com/GuyBaele/SpreadGL. Supplementary data are available at Bioinformatics online.

BICEP: Bayesian inference for rare genomic variant causality evaluation in pedigrees.

Next-generation sequencing is widely applied to the investigation of pedigree data for gene discovery. However, identifying plausible disease-causing variants within a robust statistical framework is challenging. Here, we introduce BICEP: a Bayesian inference tool for rare variant causality evaluation in pedigree-based cohorts. BICEP calculates the posterior odds that a genomic variant is causal for a phenotype based on the variant cosegregation as well as a priori evidence such as deleteriousness and functional consequence. BICEP can correctly identify causal variants for phenotypes with both Mendelian and complex genetic architectures, outperforming existing methodologies. Additionally, BICEP can correctly down-weight common variants that are unlikely to be involved in phenotypic liability in the context of a pedigree, even if they have reasonable cosegregation patterns. The output metrics from BICEP allow for the quantitative comparison of variant causality within and across pedigrees, which is not possible with existing approaches.

Statistical properties of partially observed integrated functional depths

AbstractIntegrated functional depths (IFDs) present a versatile toolbox of methods introducing notions of ordering, quantiles, and rankings into a functional data analysis context. They provide fundamental tools for nonparametric inference of infinite-dimensional data. Recently, the literature has extended IFDs to address the challenges posed by partial observability of functional data, commonly encountered in practice. That resulted in the development of partially observed integrated functional depths (POIFDs). POIFDs have demonstrated good empirical results in simulated experiments and real problems. However, there are still no theoretical results in line with the state of the art of IFDs. This article addresses this gap by providing theoretical support for POIFDs, including (i) uniform consistency of their sample versions, (ii) weak continuity with respect to the underlying probability measure, and (iii) uniform consistency for discretely observed functional data. Finally, we present a sensitivity analysis that evaluates how our theoretical results are affected by violations of the main assumptions.

Genome-wide local ancestry and the functional consequences of admixture in African and European cattle populations.

Bos taurus (taurine) and Bos indicus (indicine) cattle diverged at least 150,000 years ago and, since that time, substantial genomic differences have evolved between the two lineages. During the last two millennia, genetic exchange in Africa has resulted in a complex tapestry of taurine-indicine ancestry, with most cattle populations exhibiting varying levels of admixture. Similarly, there are several Southern European cattle populations that also show evidence for historical gene flow from indicine cattle, the highest levels of which are found in the Central Italian White breeds. Here we use two different software tools (MOSAIC and ELAI) for local ancestry inference (LAI) with genome-wide high- and low-density SNP array data sets in hybrid African and residually admixed Southern European cattle populations and obtained broadly similar results despite critical differences in the two LAI methodologies used. Our analyses identified genomic regions with elevated levels of retained or introgressed ancestry from the African taurine, European taurine, and Asian indicine lineages. Functional enrichment of genes underlying these ancestry peaks highlighted biological processes relating to immunobiology and olfaction, some of which may relate to differing susceptibilities to infectious diseases, including bovine tuberculosis, East Coast fever, and tropical theileriosis. Notably, for retained African taurine ancestry in admixed trypanotolerant cattle we observed enrichment of genes associated with haemoglobin and oxygen transport. This may reflect positive selection of genomic variants that enhance control of severe anaemia, a debilitating feature of trypanosomiasis disease, which severely constrains cattle agriculture across much of sub-Saharan Africa.

Integrating Contact Tracing Data to Enhance Outbreak Phylodynamic Inference: A Deep Learning Approach.

Phylodynamics is central to understanding infectious disease dynamics through the integration of genomic and epidemiological data. Despite advancements, including the application of deep learning to overcome computational limitations, significant challenges persist due to data inadequacies and statistical unidentifiability of key parameters. These issues are particularly pronounced in poorly resolved phylogenies, commonly observed in outbreaks such as SARS-CoV-2. In this study, we conducted a thorough evaluation of PhyloDeep, a deep learning inference tool for phylodynamics, assessing its performance on poorly resolved phylogenies. Our findings reveal the limited predictive accuracy of PhyloDeep (and other state-of-the-art approaches) in these scenarios. However, models trained on poorly resolved, realistically simulated trees demonstrate improved predictive power, despite not being infallible, especially in scenarios with superspreading dynamics, whose parameters are challenging to capture accurately. Notably, we observe markedly improved performance through the integration of minimal contact tracing data, which refines poorly resolved trees. Applying this approach to a sample of SARS-CoV-2 sequences partially matched to contact tracing from Hong Kong yields informative estimates of superspreading potential, extending beyond the scope of contact tracing data alone. Our findings demonstrate the potential for enhancing phylodynamic analysis through complementary data integration, ultimately increasing the precision of epidemiological predictions crucial for public health decision-making and outbreak control.

Local signal detection on irregular domains with generalized varying coefficient models

In spatial analysis, it is essential to understand and quantify spatial or temporal heterogeneity. This paper focuses on the generalized spatially varying coefficient model (GSVCM), a powerful framework to accommodate spatial heterogeneity by allowing regression coefficients to vary in a given spatial domain. We propose a penalized bivariate spline method for detecting local signals in GSVCM. The key idea is to use bivariate splines defined on triangulation to approximate nonparametric varying coefficient functions and impose a local penalty on L 2 norms of spline coefficients for each triangle to identify null regions of zero effects. Moreover, we develop model confidence regions as the inference tool to quantify the uncertainty of the estimated null regions. Our method partitions the region of interest using triangulation and efficiently approximates irregular domains. In addition, we propose an efficient algorithm to obtain the proposed estimator using the local quadratic approximation. We also establish the consistency of estimated nonparametric coefficient functions and the estimated null regions. The numerical performance of the proposed method is evaluated in both simulation cases and real data analysis.

Sparsified simultaneous confidence intervals for high-dimensional linear models

AbstractStatistical inference of the high-dimensional regression coefficients is challenging because the uncertainty introduced by the model selection procedure is hard to account for. Currently, the inference of the model and the inference of the coefficients are separately sought. A critical question remains unsettled; that is, is it possible to embed the inference of the model into the simultaneous inference of the coefficients? If so, then how to properly design a simultaneous inference tool with desired properties? To this end, we propose a notion of simultaneous confidence intervals called the sparsified simultaneous confidence intervals (SSCI). Our intervals are sparse in the sense that some of the intervals’ upper and lower bounds are shrunken to zero (i.e., [0, 0]), indicating the unimportance of the corresponding covariates. These covariates should be excluded from the final model. The rest of the intervals, either containing zero (e.g., $$[-1,1]$$ [ - 1 , 1 ] or [0, 1]) or not containing zero (e.g., [2, 3]), indicate the plausible and significant covariates, respectively. The SSCI intuitively suggests a lower-bound model with significant covariates only and an upper-bound model with plausible and significant covariates. The proposed method can be coupled with various selection procedures, making it ideal for comparing their uncertainty. For the proposed method, we establish desirable asymptotic properties, develop intuitive graphical tools for visualization, and justify its superior performance through simulation and real data analysis.

Tree Sequences as a General-Purpose Tool for Population Genetic Inference.

As population genetic data increase in size, new methods have been developed to store genetic information in efficient ways, such as tree sequences. These data structures are computationally and storage efficient but are not interchangeable with existing data structures used for many population genetic inference methodologies such as the use of convolutional neural networks applied to population genetic alignments. To better utilize these new data structures, we propose and implement a graph convolutional network to directly learn from tree sequence topology and node data, allowing for the use of neural network applications without an intermediate step of converting tree sequences to population genetic alignment format. We then compare our approach to standard convolutional neural network approaches on a set of previously defined benchmarking tasks including recombination rate estimation, positive selection detection, introgression detection, and demographic model parameter inference. We show that tree sequences can be directly learned from using a graph convolutional network approach and can be used to perform well on these common population genetic inference tasks with accuracies roughly matching or even exceeding that of a convolutional neural network-based method. As tree sequences become more widely used in population genetic research, we foresee developments and optimizations of this work to provide a foundation for population genetic inference moving forward.

Hidden or Inferred: Fair Learning-To-Rank With Unknown Demographics

As learning-to-rank models are increasingly deployed for decision-making in areas with profound life implications, the FairML community has been developing fair learning-to-rank (LTR) models. These models rely on the availability of sensitive demographic features such as race or sex. However, in practice, regulatory obstacles and privacy concerns protect this data from collection and use. As a result, practitioners may either need to promote fairness despite the absence of these features or turn to demographic inference tools to attempt to infer them. Given that these tools are fallible, this paper aims to further understand how errors in demographic inference impact the fairness performance of popular fair LTR strategies. In which cases would it be better to keep such demographic attributes hidden from models versus infer them? We examine a spectrum of fair LTR strategies ranging from fair LTR with and without demographic features hidden versus inferred to fairness-unaware LTR followed by fair re-ranking. We conduct a controlled empirical investigation modeling different levels of inference errors by systematically perturbing the inferred sensitive attribute. We also perform three case studies with real-world datasets and popular open-source inference methods. Our findings reveal that as inference noise grows, LTR-based methods that incorporate fairness considerations into the learning process may increase bias. In contrast, fair re-ranking strategies are more robust to inference errors. All source code, data, and experimental artifacts of our experimental study are available here: https://github.com/sewen007/hoiltr.git

Apollo: A comprehensive GPU-powered within-host simulator for viral evolution and infection dynamics across population, tissue, and cell.

Modern sequencing instruments bring unprecedented opportunity to study within-host viral evolution in conjunction with viral transmissions between hosts. However, no computational simulators are available to assist the characterization of within-host dynamics. This limits our ability to interpret epidemiological predictions incorporating within-host evolution and to validate computational inference tools. To fill this need we developed Apollo, a GPU-accelerated, out-of-core tool for within-host simulation of viral evolution and infection dynamics across population, tissue, and cellular levels. Apollo is scalable to hundreds of millions of viral genomes and can handle complex demographic and population genetic models. Apollo can replicate real within-host viral evolution; accurately recapturing observed viral sequences from an HIV cohort derived from initial population-genetic configurations. For practical applications, using Apollo-simulated viral genomes and transmission networks, we validated and uncovered the limitations of a widely used viral transmission inference tool.

QuAC: Quick Attribute-Centric Type Inference for Python

Python’s dynamic typing facilitates rapid prototyping and underlies its popularity in many domains. However, dynamic typing reduces the power of many static checking and bug-finding tools. Python type annotations can make these tools more useful. Type inference tools aim to reduce developers’ burden of adding them. However, existing type inference tools struggle to support dynamic features, infer correct types (especially container type parameters and non-builtin types), and run in reasonable time. Inspired by Python’s duck typing, where the attributes accessed on Python expressions characterize their implicit interfaces, we propose QuAC (Quick Attribute-Centric Type Inference for Python). At its core, QuAC collects attribute sets for Python expressions and leverages information retrieval techniques to predict classes from these attribute sets. It also recursively predicts container type parameters. We evaluate QuAC’s performance on popular Python projects. Compared to state-of-the-art non-LLM baselines, QuAC predicts types with high accuracy complementary to those predicted by the baselines while not sacrificing coverage. It also demonstrates clear advantages in predicting container type parameters and non-builtin types and reduces run times. Furthermore, QuAC is nearly two orders of magnitude faster than an LLM-based method while covering nearly half of its errorless non-trivial type predictions. It is also significantly more consistent at predicting container type parameters and non-builtin types than the LLM-based method, regardless of whether the project has ground-truth type annotations.

Tree sequences as a general-purpose tool for population genetic inference.

As population genetics data increases in size new methods have been developed to store genetic information in efficient ways, such as tree sequences. These data structures are computationally and storage efficient, but are not interchangeable with existing data structures used for many population genetic inference methodologies such as the use of convolutional neural networks (CNNs) applied to population genetic alignments. To better utilize these new data structures we propose and implement a graph convolutional network (GCN) to directly learn from tree sequence topology and node data, allowing for the use of neural network applications without an intermediate step of converting tree sequences to population genetic alignment format. We then compare our approach to standard CNN approaches on a set of previously defined benchmarking tasks including recombination rate estimation, positive selection detection, introgression detection, and demographic model parameter inference. We show that tree sequences can be directly learned from using a GCN approach and can be used to perform well on these common population genetics inference tasks with accuracies roughly matching or even exceeding that of a CNN-based method. As tree sequences become more widely used in population genetics research we foresee developments and optimizations of this work to provide a foundation for population genetics inference moving forward.

Causal effect estimation in survival analysis with high dimensional confounders.

With the ever advancing of modern technologies, it has become increasingly common that the number of collected confounders exceeds the number of subjects in a data set. However, matching based methods for estimating causal treatment effect in their original forms are not capable of handling high-dimensional confounders, and their various modified versions lack statistical support and valid inference tools. In this article, we propose a new approach for estimating causal treatment effect, defined as the difference of the restricted mean survival time (RMST) under different treatments in high-dimensional setting for survival data. We combine the factor model and the sufficient dimension reduction techniques to construct propensity score and prognostic score. Based on these scores, we develop a kernel based doubly robust estimator of the RMST difference. We demonstrate its link to matching and establish the consistency and asymptotic normality of the estimator. We illustrate our method by analyzing a dataset from a study aimed at comparing the effects of two alternative treatments on the RMST of patients with diffuse large B cell lymphoma.

A Novel Method for Nonparametric Statistical Inference for Niche Overlap in Multiple Species.

The understanding of species interactions and ecosystem dynamics hinges upon the study of ecological niches. Quantifying the overlap of Hutchinsonian-niches has garnered significant attention, with many recent publications addressing the issue. Prior work on estimating niche overlap often did not provide confidence intervals or assumed multivariate normality, seriously limiting applications in ecology, and biodiversity research. This paper extends a nonparametric approach, previously applied to the two-species case, to multiple species. For estimation, a consistent plug-in estimator based on rank sums is proposed and its asymptotic distribution is derived under weak conditions. The novel methodology is then applied to a study comparing the ecological niches of the Eurasian eagle owl, common buzzard, and red kite. These species share a habitat in Central Europe but exhibit distinct population trends. The analysis explores their breeding habitat preferences, considering the intricate competition dynamics and utilizing the nonparametric approach to niche overlap estimation. Our proposed method provides a valuable inferential tool for the quantitative evaluation of differences and overlap between niches.