遇见数据集

Supplementary Material for the Citation Network in 'Tracing the impact of electrophysiological studies of magnocellular neurons'

收藏
Zenodo2026-05-14 更新2026-05-29 收录
官方服务:

资源简介:

IntroductionHere, we describe the data collection and citation network dataset for ‘Tracing the impact of electrophysiological studies of magnocellular neurons’. The aim was to reconstruct: (i) the immediate citation neighbourhood of Yagi (1966); (ii) the subset of papers that directly implement Yagi’s technique; and (iii) the broader downstream literature citing those technique-using papers, before applying a series of filters to obtain a network for clustering and interpretation. The full methodology is reported in the manuscript; here we walk the reader through the main steps used to construct the dataset and resulting network dataset. Data Collection• We first retrieved all direct citations to Yagi (1966) using Web of Science’s (WoS) Cited Reference Search (81 citing items). We retained only records classified as journal paper by WoS (i.e., excluding reviews, abstracts, book chapters etc.), leaving 62 papers.• We then added in non-indexed citing papers. We added two papers that cite Yagi (1966) but are not indexed in WoS, Sundsten et al. (1970) and Novin et al. (1970), bringing the total to 64 journal papers. Because these two papers were not available as WoS records, their reference lists were manually coded and inserted into the citation data so that their data were represented.• After reading these 64 papers, we identified 31 as directly using Yagi’s technique (including Sundsten and Novin). • We then searched for each of the 31 seeds in WoS and retrieved all citing journal papers with the same filters as previously. In total, this seed set was cited by 1,150 journal papers.• We then pooled the seed-citation set with Yagi’s direct citation set (and deduplicated), resulting in 1,191 unique papers including Yagi (1966) and the 31 seed papers.• We then converted the pooled records into a directed citation network (nodes = papers; directed edges = citation links), containing 25,456 nodes connected by 48,029 citation links (including both retrieved papers and cited references).• To reduce the influence of review-like documents, which can be misclassified by WoS, we removed any retrieved source with >100 references. We also removed any referenced works that were cited only by the removed sources. This reduced the network to 21,256 nodes and 42,097 citation links, and reduced the number of retrieved papers to 1,165 connected together by 8,374 citation links.• For the final analysis reported in the paper, we restricted to the directed citation network among retrieved papers only (excluding links to non-retrieved referenced items) and set the degree inclusion criteria to ≥5, producing the final network of 853 nodes and 7,627 citation links. We clustered this via modularity maximisation via the Leiden algorithm (Traag et al. 2019), producing 7 clusters with modularity Q = 0.47. It is this clustered final network that we report here. Network Dataset Description‘Tracing_Node_List.csv’ is a comma-seperated values file that contains identifying information for the 853 papers included in the our study. The columns refer to:- Id, the node identification number (id)- Label, the reference string of the paper- wos_ut, Web of Science accession number- pmid, the pubmed id, if available- doi, the DOI for the document, if available- authors, all named authors separated by ; in authorship order- title, the title of the paper- publication_year, the year of publication- journal, the journal of publication - wos_total_citations, total citation from Web of Science Core Collection as of January 2026- indegree, total within network citations to a paper from other papers in this network- outdegree, total within network references from a paper to other papers in this network- degree, combined total of within network citations and references of a paper- cluster, the cluster id for the cluster membership discussed in the manuscript (Figure 1) ‘Tracing_Edge_List.csv’ is a comma-separated values file that contains 7,627 edges (citation links) between 853 nodes (documents). The columns refer to:- Source, node id of the citing document- Target, node id of the cited document References cited in this note: Yagi, K., Azuma, T., Matsuda, K., 1966. Neurosecretory cell: capable of conducting impulse in rats. Science 154, 778-779. Sundsten, J.W., Novin, D., Cross, B.A., 1970. Identification and distribution of paraventricular units excited by stimulation of the neural lobe of the hypophysis. Exp Neurol 26, 316-329. Novin, D., Sundsten, J.W., Cross, B.A., 1970. Some properties of antidromically activated units in the paraventricular nucleus of the hypothalamus. Exp Neurol. 26, 330-341 Traag, V.A., Waltman, L., van Eck, N.J., 2019. From Louvain to Leiden: guaranteeing well-connected communities. Sci Rep 9, 5233.

提供机构:
Zenodo
创建时间:
2026-05-14
二维码
社区交流群
二维码
科研交流群
商业服务