ILANIT 2023

How can Core Facilities improve replicability and standardization: Meta-analysis of science, CatWalk XT system output as an example

Lior Bikovski 2,3 Gilad Mass 1
1Sackler Interdepartmental Core Facility, Sackler School of Medicine, Tel Aviv University, Israel
2Myers Neuro-Behavioral Core Facility, Sackler Faculty of Medicine, Tel Aviv University, Israel
3School of Behavioral Sciences, Netanya Academic College, Israel

In the last two decades the number of academic core facilities is rising. Core Facility is an expert research unit that allow a university to pull resources and develop an expert research unit (e.g. behavior, imaging, and microscopy) for many academic users. Thiess facilities maintain top edge equipment, and develop expert personal over the years. Additional benefit of the core facility is that it can improve standardization by teaching the same methods and protocols to different labs, and via the expert personal keeping mistakes to a minimum.
In addition, it is our belief, that core facilities could have another important benefit, which could have a significant effect on understanding raw data, and improve replicability in the long run. As core facilities are a hub that information passes through and is some cases even stored in, a simple meta-analysis method of stored data could provide a powerful tool into in-depth understanding of each one of the user’s raw data.
In this abstract we present an example of our belief by performing a simple meta-analysis of published information from the Noldus CatWalk XT Automated Gait Analysis system that has been commercially available since 2006, with over 350 systems installed worldwide, and used in over 2000 publications, making the CatWalk one of the most widely implemented automated gait analysis systems today.
Focusing on seven different mouse models, we reviewed over 100 papers (between the years 2017 and 2022). The result show that once a collective analysis is done, new insights emerge that could assist novel and expert users of the system to minimize misunderstanding of raw data, and create an in-depth understanding of the data.