How Scientific Literature Restrictions Affect Future Research: the Case of Nerve Cells

Abstract

NeuroMorpho.Org is an open-source online repository that helps the scientific community by freely providing neuronal and glial reconstructions to foster brain research. The process to identify and share these data starts with literature search and evaluation. The evaluation phase combines machine learning classification and human annotation to sort the articles possibly containing data of interest as positive with high confidence, positive with low confidence, negative, or inaccessible. Here we perform a comparative analysis of all publications evaluated in 2020 and 2021 and report that the number of inaccessible papers has dramatically increased in 2021. During 2020, 13.6% articles were evaluated as positive, 83.7% as negative, and 2.7% as inaccessible. During 2021, 11.9% were positive, 73.5% as negative, and 14.6% as inaccessible. The tremendous jump in inaccessible papers from 2020 to 2021 may be due to the decision of George Mason University not to fully renew the Elsevier license. To further explore this possibility, we broke down the number of inaccessible papers. In 2021, almost all inaccessible papers (94.3%) were published by Elsevier against fewer than two-thirds (62.7%) in 2020. In conclusion, literature evaluation is a key step in the NeuroMorpho.Org pipeline, enabling and accelerating the collection on neuronal and glia reconstructions. Literature restrictions affect the number of papers that can be evaluated and, consequently, the amount of shareable data.