Cargando…

Eliminating accidental deviations to minimize generalization error and maximize replicability: Applications in connectomics and genomics

Replicability, the ability to replicate scientific findings, is a prerequisite for scientific discovery and clinical utility. Troublingly, we are in the midst of a replicability crisis. A key to replicability is that multiple measurements of the same item (e.g., experimental sample or clinical parti...

Descripción completa

Detalles Bibliográficos
Autores principales: Bridgeford, Eric W., Wang, Shangsi, Wang, Zeyi, Xu, Ting, Craddock, Cameron, Dey, Jayanta, Kiar, Gregory, Gray-Roncal, William, Colantuoni, Carlo, Douville, Christopher, Noble, Stephanie, Priebe, Carey E., Caffo, Brian, Milham, Michael, Zuo, Xi-Nian, Vogelstein, Joshua T.
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Public Library of Science 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8500408/
https://www.ncbi.nlm.nih.gov/pubmed/34529652
http://dx.doi.org/10.1371/journal.pcbi.1009279
_version_ 1784580439835410432
author Bridgeford, Eric W.
Wang, Shangsi
Wang, Zeyi
Xu, Ting
Craddock, Cameron
Dey, Jayanta
Kiar, Gregory
Gray-Roncal, William
Colantuoni, Carlo
Douville, Christopher
Noble, Stephanie
Priebe, Carey E.
Caffo, Brian
Milham, Michael
Zuo, Xi-Nian
Vogelstein, Joshua T.
author_facet Bridgeford, Eric W.
Wang, Shangsi
Wang, Zeyi
Xu, Ting
Craddock, Cameron
Dey, Jayanta
Kiar, Gregory
Gray-Roncal, William
Colantuoni, Carlo
Douville, Christopher
Noble, Stephanie
Priebe, Carey E.
Caffo, Brian
Milham, Michael
Zuo, Xi-Nian
Vogelstein, Joshua T.
author_sort Bridgeford, Eric W.
collection PubMed
description Replicability, the ability to replicate scientific findings, is a prerequisite for scientific discovery and clinical utility. Troublingly, we are in the midst of a replicability crisis. A key to replicability is that multiple measurements of the same item (e.g., experimental sample or clinical participant) under fixed experimental constraints are relatively similar to one another. Thus, statistics that quantify the relative contributions of accidental deviations—such as measurement error—as compared to systematic deviations—such as individual differences—are critical. We demonstrate that existing replicability statistics, such as intra-class correlation coefficient and fingerprinting, fail to adequately differentiate between accidental and systematic deviations in very simple settings. We therefore propose a novel statistic, discriminability, which quantifies the degree to which an individual’s samples are relatively similar to one another, without restricting the data to be univariate, Gaussian, or even Euclidean. Using this statistic, we introduce the possibility of optimizing experimental design via increasing discriminability and prove that optimizing discriminability improves performance bounds in subsequent inference tasks. In extensive simulated and real datasets (focusing on brain imaging and demonstrating on genomics), only optimizing data discriminability improves performance on all subsequent inference tasks for each dataset. We therefore suggest that designing experiments and analyses to optimize discriminability may be a crucial step in solving the replicability crisis, and more generally, mitigating accidental measurement error.
format Online
Article
Text
id pubmed-8500408
institution National Center for Biotechnology Information
language English
publishDate 2021
publisher Public Library of Science
record_format MEDLINE/PubMed
spelling pubmed-85004082021-10-09 Eliminating accidental deviations to minimize generalization error and maximize replicability: Applications in connectomics and genomics Bridgeford, Eric W. Wang, Shangsi Wang, Zeyi Xu, Ting Craddock, Cameron Dey, Jayanta Kiar, Gregory Gray-Roncal, William Colantuoni, Carlo Douville, Christopher Noble, Stephanie Priebe, Carey E. Caffo, Brian Milham, Michael Zuo, Xi-Nian Vogelstein, Joshua T. PLoS Comput Biol Research Article Replicability, the ability to replicate scientific findings, is a prerequisite for scientific discovery and clinical utility. Troublingly, we are in the midst of a replicability crisis. A key to replicability is that multiple measurements of the same item (e.g., experimental sample or clinical participant) under fixed experimental constraints are relatively similar to one another. Thus, statistics that quantify the relative contributions of accidental deviations—such as measurement error—as compared to systematic deviations—such as individual differences—are critical. We demonstrate that existing replicability statistics, such as intra-class correlation coefficient and fingerprinting, fail to adequately differentiate between accidental and systematic deviations in very simple settings. We therefore propose a novel statistic, discriminability, which quantifies the degree to which an individual’s samples are relatively similar to one another, without restricting the data to be univariate, Gaussian, or even Euclidean. Using this statistic, we introduce the possibility of optimizing experimental design via increasing discriminability and prove that optimizing discriminability improves performance bounds in subsequent inference tasks. In extensive simulated and real datasets (focusing on brain imaging and demonstrating on genomics), only optimizing data discriminability improves performance on all subsequent inference tasks for each dataset. We therefore suggest that designing experiments and analyses to optimize discriminability may be a crucial step in solving the replicability crisis, and more generally, mitigating accidental measurement error. Public Library of Science 2021-09-16 /pmc/articles/PMC8500408/ /pubmed/34529652 http://dx.doi.org/10.1371/journal.pcbi.1009279 Text en © 2021 Bridgeford et al https://creativecommons.org/licenses/by/4.0/This is an open access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/) , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
spellingShingle Research Article
Bridgeford, Eric W.
Wang, Shangsi
Wang, Zeyi
Xu, Ting
Craddock, Cameron
Dey, Jayanta
Kiar, Gregory
Gray-Roncal, William
Colantuoni, Carlo
Douville, Christopher
Noble, Stephanie
Priebe, Carey E.
Caffo, Brian
Milham, Michael
Zuo, Xi-Nian
Vogelstein, Joshua T.
Eliminating accidental deviations to minimize generalization error and maximize replicability: Applications in connectomics and genomics
title Eliminating accidental deviations to minimize generalization error and maximize replicability: Applications in connectomics and genomics
title_full Eliminating accidental deviations to minimize generalization error and maximize replicability: Applications in connectomics and genomics
title_fullStr Eliminating accidental deviations to minimize generalization error and maximize replicability: Applications in connectomics and genomics
title_full_unstemmed Eliminating accidental deviations to minimize generalization error and maximize replicability: Applications in connectomics and genomics
title_short Eliminating accidental deviations to minimize generalization error and maximize replicability: Applications in connectomics and genomics
title_sort eliminating accidental deviations to minimize generalization error and maximize replicability: applications in connectomics and genomics
topic Research Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8500408/
https://www.ncbi.nlm.nih.gov/pubmed/34529652
http://dx.doi.org/10.1371/journal.pcbi.1009279
work_keys_str_mv AT bridgefordericw eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT wangshangsi eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT wangzeyi eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT xuting eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT craddockcameron eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT deyjayanta eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT kiargregory eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT grayroncalwilliam eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT colantuonicarlo eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT douvillechristopher eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT noblestephanie eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT priebecareye eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT caffobrian eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT milhammichael eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT zuoxinian eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics
AT vogelsteinjoshuat eliminatingaccidentaldeviationstominimizegeneralizationerrorandmaximizereplicabilityapplicationsinconnectomicsandgenomics