Dear AMP-AD team, First of all, thanks again for making this great resource available using this platform. I'm not sure if this is a question that I should ask to the data access committee directly, or whether I can post it here. However, I think this question is potentially relevant for all users of this resource, so that's why I'm posting it here. Our team has recently finished a manuscript where we've combined AMP-AD RNA-seq data with multiple other brain RNA-seq datasets. We have placed a preprint on bioRxiv, and since then we are receiving e-mails from other researchers asking if we can make certain data available to them. The data access agreement clearly states that sharing of individual-level data is prohibited, and on the basis of this, we have sofar denied such requests. Other brain-related datasets such as GTEx and PsychENCODE also have the same provisions in their data access agreements, where it is prohibited to share individual level data. However, both studies publish gene expression and splicing matrices on their repository websites (e.g. on http://resource.psychencode.org/). These matrices contain a value per individual per gene or splice event, and could thus, in a strict sense, be considered individual-level data. However, since these matrices are available for download without providing registration details, this would suggest that gene expression and splicing matrices are actually not considered to be individual-level data by these studies. Upon contacting the GTEx data access committee at the NIH, this indeed seems to be the case. My question is therefore, specifically for AMP-AD datasets: does the AMP-AD consortium consider splicing or gene expression levels for an individual to be individual-level data? Would we be allowed to share such matrices after publication of our work? With kind regards, Harm-Jan

Created by Harm-Jan Westra HarmJanWestra
Dear Will, Thanks for the quick and clear answer. In our case, we realigned and re-quantified the AMP-AD samples, and processed them equally to other included datasets, in order to remove any variability that differences in alignment might have produced. Nevertheless, I'm happy that I asked. The distinction on what data can and cannot be shared is not always clear from data access agreements. I guess the conclusion is that the data cannot be shared without consent from the AMP-AD consortium. Thanks again, Harm-Jan
Hi @HarmJanWestra, Thanks for your inquiry, and it's great to hear that the data is being put to good use! Unfortunately, regardless of individual-level distinction, processed gene counts shared on the knowledge portal still fall under the same data-use restrictions as other data structures and should only be shared with users who have completed the appropriate data use certificate. We definitely want these datasets to be accessed by a diverse user-base, so I would encourage them to create a Synapse account to gain access directly, and we are happy to field questions regarding on-boarding or data use agreements. As a follow up question, were these matrix files downloaded directly from the AD Knowledge portal/Synapse, or did you process the raw data to produce the matrices? Best, Will

Are AMP-AD gene or splicing matrices considered individual-level data? page is loading…