search

neurorestore / Augur

Cell type prioritization in single-cell data
MIT License
94 stars 10 forks source link

single cell datasets #1

Open elimereu opened 4 years ago

elimereu commented 4 years ago

Hi,

I would like to test your tool in the single cell datasets you used in the paper. If you don't mind, could you provide the r objects with annotations? Thank you in advance!

skinnider commented 4 years ago

Hi Elisabetta, happy to help. In the paper, we analyzed 29 different datasets. Is there a particular one I can help you find? Are you just looking for a test dataset to play around with? Mike

elimereu commented 4 years ago

Hi Mike, thanks for your quick reply! I'm interested in many, so that's why I was wondering if you can directly provide me the r objects of your already analyzed datasets (I don't know if you also have UMAPs for example). If this is not an option, I would say I'm particularly interested in these 2 datasets: Jaitin et al and Cheng et al. Thanks a lot in advance! Bests, Elisabetta

skinnider commented 4 years ago

Hi Elisabetta - unfortunately I’m not sure whether I can provide those two datasets, because in both cases some or all of the data was not obtained from public databases such as GEO, but was provided directly by the authors (counts and metadata in the case of Cheng et al., metadata only in the case of Jaitin et al.). I can certainly ask the authors for their permission to re-distribute the data they’ve provided, though. I can also go through the remaining datasets and try to collect the ones where all the data is already public, although this will take a bit longer. Let me know what you’d prefer. In the meantime, I’ve attached one R object (for the Kang et al. 2018 dataset) in the hopes it might be useful for you to get started: https://www.dropbox.com/s/ln0owhuo3t4ry5n/GSE96583.rds?dl=0

elimereu commented 4 years ago

Hi Mike! I'm sorry for the long silence... Oh! This is a pity I can have those data.. Yes, it would be good if you can ask their permission! Thanks a lot for the data you are sharing, I will try with it..

skinnider commented 4 years ago

Hi again Elisabetta - we are working on assembling all of the preprocessed, public datasets in a single repository of Seurat objects. We hope to have this done in the next ~week or so - I will update you when it’s public. For the datasets where key data was obtained directly from the authors though, these are in a bit more of a grey area and I don’t think we can release them without permission. If there is a specific one of those you are looking for, let me know and I can inquire with the authors about re-distributing it.

elimereu commented 4 years ago

Hi Mike,

Thanks! I will wait for your repository then.. ;)

Il giorno 5 giu 2020, alle ore 18:50, Michael Skinnider notifications@github.com ha scritto:

Hi again Elisabetta - we are working on assembling all of the preprocessed, public datasets in a single repository of Seurat objects. We hope to have this done in the next ~week or so - I will update you when it’s public. For the datasets where key data was obtained directly from the authors though, these are in a bit more of a grey area and I don’t think we can release them without permission. If there is a specific one of those you are looking for, let me know and I can inquire with the authors about re-distributing it.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/neurorestore/Augur/issues/1#issuecomment-639625602, or unsubscribe https://github.com/notifications/unsubscribe-auth/AE5OXSY43NVOWWYP6WD5AKDRVEO5ZANCNFSM4KQ2SCCQ.

M0hammadL commented 3 years ago

hi Mikee, is there any update on seurat object repo? to put all the datasets you have used in the paper? How can you reproduce the results if you can't share the data?

jordansquair commented 3 years ago

We are working on providing this repository. However, to provide all the datasets used in the manuscript is complicated since some were provided directly from the authors. Which dataset @M0hammadL did you particularly have in mind? If it is one of the ones that can be reconstructed from public data you can certainly preprocess it yourself to reproduce the results.