Resumen: The aim of this paper is to establish the similarities and differences between the way of collaboration and the production of researchers when dealing with publications or with the development of projects and whether the collaboration patterns change across disciplines. We have studied the networks of researchers formed through the collaborations in papers or in projects in a research institution (the University of Zaragoza) and we have analyzed a series of individual and global magnitudes. As a general result, we have observed that the laws governing the individual productivity are similar for the cases of publications and projects but, however, the behavior is different when analyzing more complex magnitudes such as the collaborations or other structural variables. We consider also the subnetworks defined by the researchers of the different disciplines and characterize their topologies and compare the corresponding collaboration patterns. Because of the general approach, we expect most of the conclusions to be applicable to other universities or research centers. Idioma: Inglés DOI: 10.1016/j.joi.2018.11.004 Año: 2019 Publicado en: Journal of Informetrics 13, 1 (2019), 64-77 ISSN: 1751-1577 Factor impacto JCR: 4.611 (2019) Categ. JCR: INFORMATION SCIENCE & LIBRARY SCIENCE rank: 9 / 87 = 0.103 (2019) - Q1 - T1 Categ. JCR: COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS rank: 16 / 109 = 0.147 (2019) - Q1 - T1 Factor impacto SCIMAGO: 2.079 - Applied Mathematics (Q1) - Computer Science Applications (Q1) - Statistics and Probability (Q1) - Management Science and Operations Research (Q1) - Modeling and Simulation (Q1) - Library and Information Sciences (Q1)