Browse
Submit Data
Databases
API
Help

Dataset Information

0 Views

0 Connections

0 Citations

0 Reanalyses

0 Downloads

Omics score: 0

Models with higher effective dimensions tend to produce more uncertain estimates.

ABSTRACT: Mathematical models are getting increasingly detailed to better predict phenomena or gain more accurate insights into the dynamics of a system of interest, even when there are no validation or training data available. Here, we show through ANOVA and statistical theory that this practice promotes fuzzier estimates because it generally increases the model's effective dimensions, i.e., the number of influential parameters and the weight of high-order interactions. By tracking the evolution of the effective dimensions and the output uncertainty at each model upgrade stage, modelers can better ponder whether the addition of detail truly matches the model's purpose and the quality of the data fed into it.

SUBMITTER: Puy A

PROVIDER: S-EPMC9581491 | biostudies-literature | 2022 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Json Xml

Publications

Models with higher effective dimensions tend to produce more uncertain estimates.

Puy Arnald A Beneventano Pierfrancesco P Levin Simon A SA Lo Piano Samuele S Portaluri Tommaso T Saltelli Andrea A

Science advances 20221019 42

Mathematical models are getting increasingly detailed to better predict phenomena or gain more accurate insights into the dynamics of a system of interest, even when there are no validation or training data available. Here, we show through ANOVA and statistical theory that this practice promotes fuzzier estimates because it generally increases the model's effective dimensions, i.e., the number of influential parameters and the weight of high-order interactions. By tracking the evolution of the e ...[more]

PMID: 36260678

Similar Datasets

Protein sites with more coevolutionary connections tend to evolve slower, while more variable protein families acquire higher coevolutionary connections.

Project description:Background: Amino acid exchanges within proteins sometimes compensate for one another and could therefore be co-evolved. It is essential to investigate the intricate relationship between the extent of coevolution and the evolutionary variability exerted at individual protein sites, as well as the whole protein. Methods: In this study, we have used a reliable set of coevolutionary connections (sites within 10Å spatial distance) and investigated their correlation with the evolutionary diversity within the respective protein sites. Results: Based on our observations, we propose an interesting hypothesis that higher numbers of coevolutionary connections are associated with lesser evolutionary variable protein sites, while higher numbers of the coevolutionary connections can be observed for a protein family that has higher evolutionary variability. Our findings also indicate that highly coevolved sites located in a solvent accessible state tend to be less evolutionary variable. This relationship reverts at the whole protein level where cytoplasmic and extracellular proteins show moderately higher anti-correlation between the number of coevolutionary connections and the average evolutionary conservation of the whole protein. Conclusions: Observations and hypothesis presented in this study provide intriguing insights towards understanding the critical relationship between coevolutionary and evolutionary changes observed within proteins. Our observations encourage further investigation to find out the reasons behind subtle variations in the relationship between coevolutionary connectivity and evolutionary diversity for proteins located at various cellular localizations and/or involved in different molecular-biological functions.

| S-EPMC5506539 | biostudies-literature

Stabilization of parameter estimates from multiexponential decay through extension into higher dimensions.

Project description:Analysis of multiexponential decay has remained a topic of active research for over 200 years. This attests to the widespread importance of this problem and to the profound difficulties in characterizing the underlying monoexponential decays. Here, we demonstrate the fundamental improvement in stability and conditioning of this classic problem through extension to a second dimension; we present statistical analysis, Monte-Carlo simulations, and experimental magnetic resonance relaxometry data to support this remarkable fact. Our results are readily generalizable to higher dimensions and provide a potential means of circumventing conventional limits on multiexponential parameter estimation.

| S-EPMC8986819 | biostudies-literature

Gender-diverse teams produce more novel and higher-impact scientific ideas.

Project description:Science's changing demographics raise new questions about research team diversity and research outcomes. We study mixed-gender research teams, examining 6.6 million papers published across the medical sciences since 2000 and establishing several core findings. First, the fraction of publications by mixed-gender teams has grown rapidly, yet mixed-gender teams continue to be underrepresented compared to the expectations of a null model. Second, despite their underrepresentation, the publications of mixed-gender teams are substantially more novel and impactful than the publications of same-gender teams of equivalent size. Third, the greater the gender balance on a team, the better the team scores on these performance measures. Fourth, these patterns generalize across medical subfields. Finally, the novelty and impact advantages seen with mixed-gender teams persist when considering numerous controls and potential related features, including fixed effects for the individual researchers, team structures, and network positioning, suggesting that a team's gender balance is an underrecognized yet powerful correlate of novel and impactful scientific discoveries.

| S-EPMC9456721 | biostudies-literature

Indirect methods produce higher estimates of fine root production and turnover rates than direct methods.

Project description:The production and turnover of fine roots play substantial roles in the biogeochemical cycles of terrestrial ecosystems. However, the disparity among the estimates of both production and turnover, particularly due to technical limitations, has been debated for several decades. Here, we conducted a meta-analysis to compare published estimates of fine root production and turnover rates derived from different methods at the same sites and at the same sampling time. On average, the estimates of fine root production and turnover rates were 87% and 124% higher, respectively, by indirect methods than by direct methods. The substantially higher fine root production and turnover estimated by indirect methods, on which most global carbon models are based, indicate the necessity of re-assessing the global carbon model predictions for atmospheric carbon sequestration in soils as a result of the production and turnover of fine roots.

| S-EPMC3494664 | biostudies-literature

N-mixture models with camera trap imagery produce accurate abundance estimates of ungulates.

Project description:Abundance estimates inform ungulate management and recovery efforts. Yet effective and affordable estimation techniques remain absent for most ungulates lacking identifiable marks and inhabiting rugged or highly vegetated terrain. Methods using N-mixture models with camera trap imagery form an appealing solution but remain unvalidated. We assess this method using populations of desert bighorn sheep (DBS; Ovis canadensis) in New Mexico, USA, plus bison (Bison bison bison) and Texas longhorn cattle (Bos taurus taurus) in Oklahoma, USA, by calculating and comparing abundance estimates to censused values. We parsed data by 3 and 7-day intervals, using images filtered or unfiltered, and collected with motion detection or timed camera settings. We employed priors informed by subject matter experts (SME) and calculated using detection-nondetection methods. Abundance estimates from filtered images captured by motion detection in 3-day intervals included the censused value across all seasons for adult DBS, rams and ewes, indicating "best practices". This "best practices" method also captured censused values for population estimates of bison (detection-nondetection) and cattle with both priors. Our assessment validates the use of N-mixture with camera trap imagery, while presenting sampling approaches, data handling procedures and model calibration to estimate wildlife population sizes more appropriately and accurately.

| S-EPMC11682081 | biostudies-literature

Higher Elevations Tend to Have Higher Proportion of Plant Species With Glandular Trichomes.

Project description:Glandular trichomes are well known to participate in plant chemical and physical defenses against herbivores, especially herbivorous insects. However, little is known about large-scale geographical patterns in glandular trichome occurrence. Herbivory pressure is thought to be higher at low elevations because of warmer and more stable climates. We therefore predicted a higher proportion of species with glandular trichomes at low elevations than at higher elevations. We compiled glandular trichome data (presence/absence) for 6,262 angiosperm species from the Hengduan Mountains (a global biodiversity hotspot in southwest China). We tested the elevational gradient (800-5,000 m a.s.l.) in the occurrence of plant species with glandular trichomes, and its correlations with biotic (occurrence of herbivorous insects) and abiotic factors, potentially shaping the elevational gradient in the occurrence of glandular trichomes. We found a significantly positive relationship between elevation and the occurrence of glandular trichomes, with the proportion of species having glandular trichomes increasing from 11.89% at 800 m a.s.l. to 17.92% at above 4,700 m. This cross-species relationship remained significant after accounting for phylogenetic relationships between species. Herbivorous insect richness peaked at mid-elevations and its association with the incidence of glandular trichomes was weak. Mean annual temperature was the most important factor associated negatively with glandular trichomes. Our results do not support the hypothesis that plant defenses decrease with increasing elevation. In contrast, a higher proportion of plant species with glandular trichome toward higher elevations is observed. Our results also highlight the importance of considering the simultaneous influences of biotic and abiotic factors in testing geographical variation in multifunctional plant defenses.

| S-EPMC8075162 | biostudies-literature

Using Random Effect Models to Produce Robust Estimates of Death Rates in COVID-19 Data.

Project description:Tracking the progress of an infectious disease is critical during a pandemic. However, the incubation period, diagnosis, and treatment most often cause uncertainties in the reporting of both cases and deaths, leading in turn to unreliable death rates. Moreover, even if the reported counts were accurate, the "crude" estimates of death rates which simply divide country-wise reported deaths by case numbers may still be poor or even non-computable in the presence of small (or zero) counts. We present a novel methodological contribution which describes the problem of analyzing COVID-19 data by two nested Poisson models: (i) an "upper model" for the cases infected by COVID-19 with an offset of population size, and (ii) a "lower" model for deaths of COVID-19 with the cases infected by COVID-19 as an offset, each equipped with their own random effect. This approach generates robustness in both the numerator as well as the denominator of the estimated death rates to the presence of small or zero counts, by "borrowing" information from other countries in the overall dataset, and guarantees positivity of both the numerator and denominator. The estimation will be carried out through non-parametric maximum likelihood which approximates the random effect distribution through a discrete mixture. An added advantage of this approach is that it allows for the detection of latent subpopulations or subgroups of countries sharing similar behavior in terms of their death rates.

| S-EPMC9690214 | biostudies-literature

Languages with more speakers tend to be harder to (machine-)learn.

Project description:Computational language models (LMs), most notably exemplified by the widespread success of OpenAI's ChatGPT chatbot, show impressive performance on a wide range of linguistic tasks, thus providing cognitive science and linguistics with a computational working model to empirically study different aspects of human language. Here, we use LMs to test the hypothesis that languages with more speakers tend to be easier to learn. In two experiments, we train several LMs-ranging from very simple n-gram models to state-of-the-art deep neural networks-on written cross-linguistic corpus data covering 1293 different languages and statistically estimate learning difficulty. Using a variety of quantitative methods and machine learning techniques to account for phylogenetic relatedness and geographical proximity of languages, we show that there is robust evidence for a relationship between learning difficulty and speaker population size. However, contrary to expectations derived from previous research, our results suggest that languages with more speakers tend to be harder to learn.

| S-EPMC10613286 | biostudies-literature

TTN truncation variants produce sarcomere-integrating proteins of uncertain functional significance.

Project description:Titin (TTN) is one of the largest and most complex proteins expressed in humans, and truncation variants are the most prevalent genetic lesion identified in individuals with dilated cardiomyopathy (DCM) or other disorders of impaired cardiac contractility. Two reports in this issue of the JCI shed light on a potential mechanism involving truncated TTN sarcomere integration and the potential for disruption of sarcomere structural integrity. Kellermayer, Tordai, and colleagues confirmed the presence of truncated TTN protein in human DCM samples. McAfee and authors developed a patient-specific TTN antibody to study truncated TTN subcellular localization and to explore its functional consequences. A "poison peptide" mechanism emerges that inspires alternative therapeutic approaches while opening new lines for inquiry, such as the role of haploinsufficiency of full-length TTN protein, mechanisms explaining sarcomere dysfunction, and explanations for variable penetrance.

| S-EPMC10786689 | biostudies-literature

Predicting global community properties from uncertain estimates of interaction strengths.

Project description:The community matrix measures the direct effect of species on each other in an ecological community. It can be used to determine whether a system is stable (returns to equilibrium after small perturbations of the population abundances), reactive (perturbations are initially amplified before damping out), and to determine the response of any individual species to perturbations of environmental parameters. However, several studies show that small errors in estimating the entries of the community matrix translate into large errors in predicting individual species responses. Here, we ask whether there are properties of complex communities one can still predict using only a crude, order-of-magnitude estimate of the community matrix entries. Using empirical data, randomly generated community matrices, and those generated by the Allometric Trophic Network model, we show that the stability and reactivity properties of systems can be predicted with good accuracy. We also provide theoretical insight into when and why our crude approximations are expected to yield an accurate description of communities. Our results indicate that even rough estimates of interaction strengths can be useful for assessing global properties of large systems.

| S-EPMC4535399 | biostudies-literature

OmicsDI is part of the ELIXIR infrastructure

OmicsDI is an Elixir interoperability service. Learn more ›

Tweets

OmicsDI Databases

PRIDE
PeptideAtlas
MassIVE
JPOST Repository
Physiome Model Repository

EGA
EVA
ENA
LINCS
PAXDB
Cell Collective

MetaboLights
Metabolomics Workbench
MetabolomeExpress
GNPS
BioModels
FAIRDOMHub

ArrayExpress
dbGaP
ExpressionAtlas
GEO
NODE

Information

Databases
Help
API
Contact us
Code on GitHub
Terms of use
Submit Data