BACKGROUND
The concept of a meta-topical brainforest is proposed, to reflect a link between collaborative research and complex ecosystems. Tropical rainforests leverage a diversity of species to capture and convert solar energy into carbon-based life, and research teams can harvest a similar benefit from a diversity of data, tools, and thought paradigms.
According to the National Institutes of Health (NIH), team science is “a collaborative and often cross-disciplinary approach to scientific inquiry that draws researchers who otherwise work independently or as co-investigators on smaller-scale projects into collaborative centers and groups” 1. Thus, team science occurs when artificial boundaries such as departments and institutions are crossed, allowing collaboration in integrated networks. Over the past two decades, the concept has received increasing attention to better understand and address global challenges 2. In 2007, Stefan Wuchty et al. examined 19.9 million research articles in the Institute for Scientific Information Web of Science database and 2.1 million patent records on multiple topics. They concluded that a team-authored paper has increased probability of being highly cited 3. The systems being formed through interdisciplinary collaborations help teams reach achievements that individual researchers are less likely to accomplish.
Kohane pointed out 4 that precision medicine in particular requires a higher level of coordination between various agencies and suggests the boundaries between research projects and clinical care institutions should be blurred to link gathered data. The exponential growth and causal interdependencies of ‘-omics’ fields dictate that expertise across disciplines is essential to making meaningful and durable contributions to the understanding of human biology.
OBJECTIVE
This brief viewpoint aims to explore the impact of cross-institution team science on the development of precision medicine. We hypothesized that international organizations with co-leaders tend to publish more impactful papers than organizations without. Using Pearson's chi-square test and the Mann-Whitney U test, we validated our hypothesis.
METHODS
Information was collected from the eHealth Catalogue of Activities developed by the nonprofit Global Alliance for Genomics and Health (GA4GH) in 2015 5. The catalog lists international genomic and clinical data-sharing initiatives, and the eHealth Task Team updated the catalog through 2017. The data on the executive leadership team and publications were obtained from the websites of these organizations. If such information was not found, additional data were acquired by directly contacting the organizations or searching on Google Scholar. The impact of papers was evaluated by their number of citations, a criterion of research quality 3.
In this paper, co-leadership means that a person holds a leadership position in different organizations concurrently. If two papers from separate organizations have at least one author in common, these two organizations are regarded as having a co-author relationship.
Nonparametric tests were performed to verify the hypothesis. We used SPSS version 22.0 (SPSSInc) and R to perform two-tailed tests with an α level of .05. The significance of the correlation between the nominal variables co-leadership and co-authorship was examined by Pearson's chi-square test of independence and expressed in a contingency table. Pearson's chi-square test of goodness of fit was adopted to evaluate whether organizations with co-leaders had a greater number of publications than organizations without, and the Mann-Whitney U test was used to examine whether the former organizations published papers that received more citations than the latter.
RESULTS
We analyzed data from 69 organizations in the catalog and found 16 pairs with co-leader relationships in 2015. Among the 374 publications from these organizations at that time, 13 pairs had co-authors. By 2017, the number of institutions in the catalog increased to 87, and there were 37 pairs with co-leadership, corresponding to 30 organizations. Information on 7,064 papers was collected, showing that 55 organizations had co-authored publications, with 436 papers in total.
A. Number of publications
The chi-square goodness of fit test suggests that the number of papers being published is strongly correlated with the category of the organization - organizations in a co-leadership network or organizations without a co-leadership (P<0.001, 2015 & 2017).
B. Quality of publications
The citation number of each paper was obtained from Google Scholar. The results of the Mann-Whitney U test indicated that the number of citations received by publications of organizations with and without co-leaders differed significantly (Z=-13.547, p<0.001, 2017). Papers from the former organizations had a higher mean rank (3603.35 for the group of papers whose publishers are in the co-leadership network, and 2702.67 for the other group), which means that the organizations with co-leaders tended to have a greater number of highly cited papers.
C. Relationship between co-leader and co-author
In the chi-square test of independence, the total sample size is the number of lines in a fully connected diagram. The results indicate that in both 2015 and 2017, organizations with co-leaders tended to publish papers together, suggesting that co-leadership will lead to co-authorship (P<0.001, 2015 & 2017).
CONCLUSIONS
These results illustrate the concept of meta-topical brainforests in precision medicine and may have broader implication: cross-enterprise cooperation plays an essential role in solving complex issues. As a field-crossing example, Sovacool suggested researchers should incorporate expertise and data from indigenous groups to address global environmental challenges 6.
One hopes the analogy persists and the extraordinary natural future-proofing mechanisms in rainforests by incorporating novel combinations of ancestral DNA coincide with similar continued diversification in research networks and widely impactful publication.
The π-index: a new indicator for assessing scientific impact
