Predictive Testing Research Articles

Transformed regression-based long-horizon predictability tests

We propose new tests for long-horizon predictability based on IVX estimation of a transformed regression which explicitly accounts for the over-lapping nature of the dependent variable in the long-horizon regression arising from temporal aggregation. To improve efficiency, we moreover incorporate the residual augmentation approach recently used in the context of short-horizon predictability testing by Demetrescu and Rodrigues (2022). Our proposed tests improve on extant tests in the literature in a number of ways. First, they allow practitioners to remain ambivalent over the strength of the persistence of the predictors. Second, they are valid under much weaker conditions on the innovations than extant long-horizon predictability tests; in particular, we allow for general forms of conditional and unconditional heteroskedasticity in the innovations, neither of which are tied to a parametric model. Third, unlike the popular Bonferroni-based methods in the literature, our proposed tests can handle multiple predictors, and can be easily implemented as either one or two-sided hypotheses tests. Monte Carlo analysis suggests that our preferred tests offer improved finite sample properties compared to the leading tests in the literature. We report results from an empirical application investigating the use of real exchange rates for predicting nominal exchange rates and inflation.

Minors at risk of von Hippel-Lindau disease: 10 years' experience of predictive genetic testing and follow-up adherence.

Von Hippel-Lindau (VHL) disease is one of the most common cancer predisposition syndromes. Penetrance is high with around 20% of children presenting detectable and curable manifestations of the disease at 15 years old. VHL predictive genetic testing (PGT) is recommended during childhood from age 5 years in France. Insufficient compliance to surveillance of VHL pathogenic variant (PV) carriers is associated with severe outcome. PGT experienced by children and their parents is probably critical in influencing future acceptance of the result and adherence to surveillance. We conducted a retrospective study on minors tested (aged 5 to 16 years old) from 2010 to 2020, in a multidisciplinary oncogenetics consultation which follows a 3-step protocol based on psychological familial support. The objectives were to assess the adherence to follow-up within the National Expert Center for inherited predispositions to renal tumors (PREDIR) network of VHL PV carriers and its benefit through tumor detection and medical interventions. A VHL PGT was carried out in 34 children. Among the 16 children diagnosed as VHL PV carriers addressed to the PREDIR network, none had discontinued surveillance after a median of 41 months. Follow-up examinations detected 11 tumors in 6 children, 4 have been surgically treated. All had a favorable outcome. Our data suggest that a specific and adapted procedure for PGT in at-risk VHL children as well as a follow-up, organized within a specialized expert network, fosters a complete adherence to the surveillance protocol and thus lead to a favorable clinical outcome.

A Power Booster Factor for Out-of-Sample Tests of Predictability

In this paper we introduce a “power booster factor” for out-of-sample tests of predictability. The relevant econometric environment is one in which the econometrician wants to compare the population Mean Squared Prediction Errors (MSPE) of two models: one big nesting model, and another smaller nested model. Although our factor can be used to improve finite sample properties of several out-of-sample tests of predictability, in this paper we focus on the widely used test developed by Clark and West (2006, 2007). Our new test multiplies the Clark and West t-statistic by a factor that should be close to one under the null hypothesis that the short nested model is the true model, but that should be greater than one under the alternative hypothesis that the big nesting model is more adequate. We use Monte Carlo simulations to explore the size and power of our approach. Our simulations reveal that the new test is well sized and powerful. In particular, it tends to be less undersized and more powerful than the test by Clark and West (2006, 2007). Although most of the gains in power are associated to size improvements, we also obtain gains in size-adjusted-power. Finally we illustrate the use of our approach when evaluating the ability that an international core inflation factor has to predict core inflation in a sample of 30 OECD economies. With our “power booster factor” more rejections of the null hypothesis are obtained, indicating a strong influence of global inflation in a selected group of these OECD countries.

Impact of BRCA1/2 cascade testing on anxiety, depression, and cancer worry levels among unaffected relatives in a multiethnic Asian cohort.

Cascade testing for families with BRCA pathogenic variants is important to identify relatives who are carriers. These relatives can benefit from appropriate risk management and preventative strategies arising from an inherited increased risk of breast, ovarian, prostate, melanoma, and pancreatic cancers. Cascade testing has the potential to enable cost-effective cancer control even in low- and middle-income settings, but few studies have hitherto evaluated the psychosocial impact of cascade testing in an Asian population, where the cultural and religious beliefs around inheritance and destiny have previously been shown to influence perception and attitudes toward screening. In this study, we evaluated the short- and long-term psychosocial impact of genetic testing among unaffected relatives of probands identified through the Malaysian Breast Cancer Genetics Study and the Malaysian Ovarian Cancer Study, using validated questionnaires (Hospital Anxiety and Depression Scale and Cancer Worry Scale) administered at baseline, and 1-month and 2-year post-disclosure of results. Of the 305 unaffected relatives from 98 independent families who were offered cascade testing, 256 (84%) completed predictive testing and family history of cancers was the only factor significantly associated with uptake of predictive testing. We found that the levels of anxiety, depression, and cancer worry among unaffected relatives decreased significantly after result disclosure and remained low 2-year post-result disclosure. Younger relatives and relatives of Malay descent had higher cancer worry at both baseline and after result disclosure compared to those of Chinese and Indian descent, whereas relatives of Indian descent and those with family history of cancers had higher anxiety and depression levels post-result disclosure. Taken together, the results from this Asian cohort highlight the differences in psychosocial needs in different communities and inform the development of culture-specific genetic counseling strategies.

Machine learning guided phase formation prediction of high entropy alloys

High entropy alloys (HEAs) have attracted intensive attention in recent years, because of their numerous structures and unusual properties. Structure prediction plays a key role in HEAs development due to the strong link between structures and properties. Thus, a new approach to rapidly predict HEAs phase formation with high accuracy has to be proposed. Here, we built a HEAs phase selection strategy based on a large as-cast dataset containing 2043 alloys data. Our dataset consists of HEAs, binary and ternary alloys. Our phase selection strategy is a combination of multi k-nearest neighbor learners with an ensemble learning method. Two new thermodynamic parameters have been proposed to improve the machine learning model’s predicting performance. Our strategy shows a surprisingly high predictability (test accuracy is 93%), and all the test accuracy values of prediction for each phase are above 97%, which means multi phases formation could be completely and detailed predicted via our phase selection strategy. Our strategy provides an alternative route of HEAs phase formation prediction that helps accelerate the development of HEAs.

Survey on implementation of molecular testing in ovarian cancer and PARP inhibitor: a national North-Eastern German Society of Gynecologic Oncology/Young Academy of Gynecologic Oncology/Arbeitsgemeinschaft Gynäkologische Onkologie intergroup analysis

BackgroundSince the introduction of poly (ADP-ribose) polymerase (PARP) inhibitors, BRCA testing has evolved as a standard management in epithelial ovarian cancer.ObjectiveTo analyze the implementation of molecular testing and PARP inhibitor...

Giant tree frogs exemplify the promise of integrating multiple types of data in phylogeographic investigations.

Hugall et al. (2022) is one of the seminal publications from the single locus era of phylogeographic research. These authors were among the first to argue that genetic data are ideally suited to test hypotheses that are ultimately derived from other sources of information. While the testing of predictions from the fossil record has long been important to molecular systematics (e.g., Donoghue et al., 1989), phylogeographic investigations into the more recent evolutionary past lack a fossil record in most focal taxa. In lieu of fossils, which were not available for the small snails that served as the focal taxon, Hugall et al. (2002) applied the (then) new technique of environmental modelling to identify regions within the species range with habitat that was predicted to be stable throughout the Holocene. They then present data that suggests that these regions correspond to the areas with high genetic diversity. Apart from the inferences about snail evolutionary history, the core argument of Hugall et al. (2002) is that consilience (i.e., agreement between inferences drawn from different sources of data) is an important goal for phylogeographic investigation. Consilience in the inferences drawn from independent types of data has a multiplicative effect; when present the researcher is likely to have more confidence in their inference than would be possible to have in an inference from any one source of data. The manuscript by Jaynes et al. (2022) is a splendid illustration of this principle.

Predictive Mutation Analysis via the Natural Language Channel in Source Code

Mutation analysis can provide valuable insights into both the system under test and its test suite. However, it is not scalable due to the cost of building and testing a large number of mutants. Predictive Mutation Testing (PMT) has been proposed to reduce the cost of mutation testing, but it can only provide statistical inference about whether a mutant will be killed or not by the entire test suite. We propose Seshat, a Predictive Mutation Analysis (PMA) technique that can accurately predict the entire kill matrix , not just the Mutation Score (MS) of the given test suite. Seshat exploits the natural language channel in code, and learns the relationship between the syntactic and semantic concepts of each test case and the mutants it can kill, from a given kill matrix. The learnt model can later be used to predict the kill matrices for subsequent versions of the program, even after both the source and test code have changed significantly. Empirical evaluation using the programs in Defects4J shows that Seshat can predict kill matrices with an average F-score of 0.83 for versions that are up to years apart. This is an improvement in F-score by 0.14 and 0.45 points over the state-of-the-art PMT technique and a simple coverage-based heuristic, respectively. Seshat also performs as well as PMT for the prediction of the MS only. When applied to a mutant-based fault localisation technique, the predicted kill matrix by Seshat is successfully used to locate faults within the top 10 position, showing its usefulness beyond prediction of MS. Once Seshat trains its model using a concrete mutation analysis, the subsequent predictions made by Seshat are on average 39 times faster than actual test-based analysis. We also show that Seshat can be successfully applied to automatically generated test cases with an experiment using EvoSuite.

Measure of predictability of a stationary two dimensionally indexed autoregressive moving-average models

In this article, we consider a measure of predictability of a random field following a stationary two-dimensionally indexed autoregressive moving-average (2-D ARMA). The purpose of this article is two-fold. First, we give a characterization of equal predictability processes by providing necessary and sufficient conditions, which highlight the role of the coefficients of the moving average (MA) and the equivalent autoregressive (EAR) representations. Second, we implement a procedure allowing us to carry out a test of equal predictability. A simulation study is presented and an application to real data of wheat yield is reported.

Cascade testing for inherited cardiac conditions: Risk perception and screening after a negative genetic test result.

First-degree relatives of a proband with an inherited cardiac condition (ICC) are offered predictive genetic testing for the pathogenic or likely pathogenic (P/LP) cardiac gene variant (CGV) to clarify their risk for the familial condition. Relatives who test negative for a familial P/LP CGV typically do not require longitudinal cardiac surveillance. To our knowledge, no previous study has investigated adjustment to risk reduction and subsequent screening practices in genotype-negative relatives from an ICC population. We thus investigated risk perception and ongoing screening practices in genotype-negative adults who received cardiac genetic counseling. Correlations between clinical and demographic variables and risk perception and screening practices were also investigated. On average, participants (n= 71) reported a perceived 19.5% lifetime risk of developing the ICC in their family, despite their negative genetic test result. The majority (54%) of participants reported having undergone cardiac screening after disclosure of their negative result. There were no significant correlations between clinical and demographic variables and risk perception or screening practices. Furthermore, risk perception was not found to impact the likelihood of cardiac screening. These findings suggest that even with comprehensive cardiac genetic counseling, a proportion of this population did not accurately comprehend or recall their cardiac disease risk. Additional interventions beyond traditional result disclosure should be explored to help genotype-negative individuals adjust to their reduction in risk for a familial ICC.

Innovation links, information diffusion, and return predictability: Evidence from China

Based on the activities of patent citation in China, a novel type of cross-firm innovation links is generated to investigate the gradual diffusion of information along the innovation chain via tests of cross-sectional return predictability. Various signals are created to represent the value of the information contained in the innovation links; these signals are demonstrated to have robust cross-predictability for stock returns in both the cross-sectional regression model and portfolio strategies. The effect of predictability is found to be stronger for stocks with high institutional ownership and analyst coverage. Considering the minimum number of steps required to establish the cross-firm linkage, innovation links are further partitioned to represent different proximity of the linked firms. It is then found that information diffuses faster across closely-linked firms than across distantly-linked firms. Sophisticated investors are found to be able to properly process the relevant information and benefit from innovation links.

Economic policy uncertainty and stock returns among OPEC members: evidence from feasible quasi-generalized least squares

We examine the predictive ability of economic policy uncertainty on stock returns of selected OPEC countries. In order to deal with certain statistical properties of the predictors, which include serial correlation, persistence, conditional heteroskedasticity, and endogeneity effects, wse utilize the Feasible Quasi-Generalized Least Squares (FQGLS) estimator in order to obtain accurate forecast estimates. As a precondition for forecast analysis, we conduct the predictability test, which shows that economic policy uncertainty is significant only for five countries, namely Kuwait, Nigeria, Saudi Arabia, United Arab Emirates, and Venezuela. Hence, we proceed with the main forecast analysis for only this set of countries. Our results are twofold. We first account for asymmetries in forecasting stock returns by comparing the forecast performance of the symmetric economic policy uncertainty-based predictive model with its asymmetric variant. On the other hand, we compare the performance of the best model from above with the standard ARFIMA model using an alternative forecast test. In both cases, we find that the asymmetric model yields the most accurate forecast returns for stock returns of the five countries. In essence, neglecting the role of asymmetries in forecasting stock returns can lead to bias results. Our findings are not only robust to different sample sizes (i.e., 50%, and 75%) and different forecast horizons (4, 8, and 12 months) but have important policy implications for policymakers and potential investors.

Clinicopathological and molecular characteristics of fumarate hydratase–deficient uterine smooth muscle tumors: a single-center study of 52 cases

The fumarate hydratase (FH) gene germline mutations cause hereditary leiomyomatosis and renal cell carcinoma syndrome (HLRCC), predisposing carriers to uterine and cutaneous leiomyomas and renal cell carcinoma. In this study, we aim to investigate morphology and the correlation between FH mutation in FH-deficient (FH-d) uterine smooth muscle tumors (uSMTs). We conducted immunohistochemical staining in 161 cases of uSMTs to detect FH deficiency. We identified 52 cases (52/161, 32%)of FH-d, including 34 leiomyomas with bizarre nuclei, 10 uSMTs of uncertain malignant potential (STUMPs), 4 cellular leiomyomas, 3 usual type leiomyomas, and 1 leiomyosarcoma. Patients with FH-d were aged 24-67 years (median, 40 years). The most common FH-d morphological features included staghorn-shaped blood vessels (87%), bizarre nuclei (81%), alveolar pattern edema (65%), macronucleoli surrounded by a halo (65%), cytoplasmic eosinophilic globules (56%), and chain-like distribution of smooth muscle cells (52%). A targeted next-generation sequence was performed in 11 of 52 FH-d tumors. Five cases (5/11, 45%) were found with FH germline mutations, including 4 leiomyomas with bizarre nuclei and 1 STUMP. The median age of patients with germline FH mutation was 30 years. The germline mutations included 3 pathogenic, 1 likely pathogenic, and 1 rare uncertain clinical significance variants. Our results revealed that FH-d uSMTs usually exhibit the distinct morphology features and high frequency of FH germline mutations. The combination of predictive morphology evaluation, FH immunotype, and molecular testing is helpful for the screening of HLRCC in uSMTs.

Forecasting effects of transport infrastructure on endangered tigers: a tool for conservation planning.

The rapid development of transport infrastructure is a major threat to endangered species worldwide. Roads and railways can increase animal mortality, fragment habitats, and exacerbate other threats to biodiversity. Predictive models that forecast the future impacts to endangered species can guide land-use planning in ways that proactively reduce the negative effects of transport infrastructure. Agent-based models are well suited for predictive scenario testing, yet their application to endangered species conservation is rare. Here, we developed a spatially explicit, agent-based model to forecast the effects of transport infrastructure on an isolated tiger (Panthera tigris) population in Nepal’s Chitwan National Park—a global biodiversity hotspot. Specifically, our model evaluated the independent and interactive effects of two mechanisms by which transport infrastructure may affect tigers: (a) increasing tiger mortality, e.g., via collisions with vehicles, and (b) depleting prey near infrastructure. We projected potential impacts on tiger population dynamics based on the: (i) existing transportation network in and near the park, and (ii) the inclusion of a proposed railway intersecting through the park’s buffer zone. Our model predicted that existing roads would kill 46 tigers over 20 years via increased mortality, and reduced the adult tiger population by 39% (133 to 81). Adding the proposed railway directly killed 10 more tigers over those 20 years; deaths that reduced the overall tiger population by 30 more individuals (81 to 51). Road-induced mortality also decreased the proportion of time a tiger occupied a given site by 5 years in the 20-year simulation. Interestingly, we found that transportation-induced depletion of prey decreased tiger occupancy by nearly 20% in sites close to roads and the railway, thereby reducing tiger exposure to transportation-induced mortality. The results of our model constitute a strong argument for taking into account prey distributions into the planning of roads and railways. Our model can promote tiger-friendly transportation development, for example, by improving Environmental Impact Assessments, identifying “no go” zones where transport infrastructure should be prohibited, and recommending alternative placement of roads and railways.

Perceptions of causal attribution and attitudes to genetic testing among people with schizophrenia and their first-degree relatives

Rapid advances in the genetics of psychiatric disorders mean that diagnostic and predictive genetic testing for schizophrenia risk may one day be a reality. This study examined how causal attributions for schizophrenia contribute to interest in a hypothetical genetic test. People with schizophrenia and first-degree relatives of people with schizophrenia were recruited through a schizophrenia research bank and mental health organisation. Semi-structured telephone interviews were conducted with 13 individuals with schizophrenia and 8 first-degree relatives. Transcripts were subjected to a qualitative analysis using the thematic analysis framework. Five themes were developed: (i) “It is like a cocktail”, with most participants aware that both genetic and environmental factors contributed to causation, and many mentioning the positive impact of genetic causal explanations; (ii) “Knowledge is power” (i.e., in favour of genetic testing); (iii) Genetic testing provides opportunities for early intervention and avoiding triggers, with participants citing a wide range of perceived benefits of genetic testing but few risks; (iv) Views on reproductive genetic testing for schizophrenia risk with a few participants viewing it as “playing God” but not necessarily being against it; and (v) “It snowballs”, whereby participants’ understanding of genetics was sophisticated with most believing that multiple rather than single genes contributed to schizophrenia. In conclusion, many individuals had a sound understanding of the role of genetic testing if it were to become available, with evidence of insight into the role of multiple genes and the contribution of other risk factors that may interact with any inherited genetic risk.

The positive effects of parents’ education level on children’s mental health in Indonesia: a result of longitudinal survey

BackgroundMental health problems are associated with decreasing the quality of various aspects of life. Cases of mental health problems (e.g., depressive symptoms) have increased continuously. Researchers predicted depression to be the first cause of world burden diseases in 2030. One of the determinant factors of mental health is parents’ education levels, but there have been contradictory research findings. The current research investigates the effect parental education attainment has on children’s mental health.MethodsWe used public data from two waves of the Indonesia Family Life Survey (IFLS) conducted in 2007 and 2014. There were 767 participants aged 15-19 years old (M = 16.80, SD = 1.37) in 2007. All participants were re-interviewed in 2014. We examined the highest level of the parents’ formal education in 2007 and the children’s mental health in 2014 to find the longitudinal effects. We used depressive symptoms and happiness as representative variables of mental health. The structural equation model (SEM) was used to examine the hypothesis, and we prioritized predictive testing over the models’ goodness of fit. We have built 12 models of combinations of children’s and parents’ sexes and different independent variables.ResultsThe hypothesis testing showed the longitudinal effects that fathers’ education in 2007 has on daughters’ depressive symptoms in 2014 (β = −.203, p < 0.01), while there were longitudinal effects from mothers’ education in 2007 on their daughters’ depressive symptoms (β = −.163, p < 0.01) and sons’ depressive symptoms (β = .096, p ≤ 0.05) in 2014. Testing the happiness models showed that fathers’ education in 2007 influenced the happiness of all of participants (β = .167, p < 0.01), including both sons (β = .206, p < 0.01) and daughters (β = 149, p < 0.01). On the contrary, no significant correlation was found between mothers’ education and children’s happiness across all three categories of participants.ConclusionsThe general results of this study showed that parents’ education levels were associated with their children’s mental health, but there are different associations found through the different combinations of children’s and parents’ sexes.

Predictive Biomarkers in Thyroid Cancer.

In molecular pathology, predictive biomarkers identify which patients are likely to respond to targeted drugs. These therapeutic agents block specific molecules directly involved in cancer growth, dedifferentiation and progression. Until few years ago, the only targeted drugs available for advanced thyroid cancer included multi-tyrosine kinase inhibitors, mainly targeting the MAPK pathway and the angiogenic signaling. The administration of these drugs does not necessarily require a molecular characterization of tumors to assess the presence of predictive alterations. However, the availability of new selective targeted drugs for thyroid cancer patients is changing the diagnostic strategies for the molecular characterization of these tumors. The search for targetable alterations can be performed directly on tumor tissue by using a variety of methodologies, depending also on the number and type of alterations to test (i.e. single nucleotide variation or gene rearrangement). Herein, a comprehensive review of the currently available targeted treatments for thyroid cancer, related predictive markers and testing methodologies is provided.

Making Antimicrobial Susceptibility Testing More Physiologically Relevant with Bicarbonate?

ABSTRACTAzithromycin is a clinically important drug for treating invasive salmonellosis despite poor activity in laboratory assays for MIC. Addition of the main buffer in blood, bicarbonate, has been proposed for more physiologically relevant and more predictive testing conditions. However, we show here that bicarbonate-triggered lowering of azithromycin MIC is entirely due to alkalization of insufficiently buffered media. In addition, bicarbonate is unlikely to be altering efflux pump activity.

Extensions to IVX methods of inference for return predictability

The contribution of this paper is threefold. First, we demonstrate that, provided either a suitable bootstrap implementation is employed or heteroskedasticity-consistent standard errors are used, the IVX-based predictability tests of Kostakis et al. (2015) retain asymptotically valid inference under the null hypothesis under considerably weaker assumptions on the innovations than are required by Kostakis et al. (2015). Second, under the same assumptions, we develop asymptotically valid bootstrap implementations of the IVX tests. Monte Carlo simulations show that the bootstrap tests deliver considerably more accurate finite sample inference than the asymptotic implementations of the tests under certain problematic parameter constellations, most notably for one-sided testing, and where multiple predictors are included. Third, we show how sub-sample implementations of the IVX approach can be used to develop asymptotically valid one-sided and two-sided tests for the presence of temporary windows of predictability.

Combining p-values for Multivariate Predictive Ability Testing

In this paper, we propose an intersection-union test for multivariate forecast accuracy based on the combination of a sequence of univariate tests. The testing framework evaluates a global null hypothesis of equal predictive ability using any number of univariate forecast accuracy tests under arbitrary dependence structures, without specifying the underlying multivariate distribution. An extensive Monte-Carlo simulation exercise shows that our proposed test has very good size and power properties under several relevant scenarios, and performs well in both low- and high-dimensional settings. We illustrate the empirical validity of our testing procedure using a large dataset of 84 daily exchange rates running from 1 January 2011 to 1 April 2021. We show that our proposed test addresses inconclusive results that often arise in practice.