Papers of interest

[seandavi]

Just for communicating about interesting papers related to multiassay experiments.

http://onlinelibrary.wiley.com/doi/10.1002/psp4.12107/full

[aedin]

Also see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5240259/

[vjcitn]

"Another disadvantage is that *MultiAssayExperiment* is being developed by a large number of contributors making package improvements slow and tedious." thank you very much

…

On Mon, Jan 30, 2017 at 3:19 PM, Aedin Culhane ***@***.***> wrote: Also see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5240259/ — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <https://github.com/waldronlab/MultiAssayExperiment/issues/165#issuecomment-276178161>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEaOwp9dw3YPlShTBkZDUUG83VrdAXA6ks5rXkXHgaJpZM4KqWGx> .

[kasperdanielhansen]

So let's figure out what they did which is good and what is less good. Best, Kasper On Mon, Jan 30, 2017 at 4:27 PM, Vince Carey <notifications@github.com> wrote:

…

"Another disadvantage is that *MultiAssayExperiment* is being developed by a large number of contributors making package improvements slow and tedious." thank you very much On Mon, Jan 30, 2017 at 3:19 PM, Aedin Culhane ***@***.***> wrote: > Also see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5240259/ > > — > You are receiving this because you are subscribed to this thread. > Reply to this email directly, view it on GitHub > <https://github.com/waldronlab/MultiAssayExperiment/issues/ 165#issuecomment-276178161>, > or mute the thread > <https://github.com/notifications/unsubscribe-auth/ AEaOwp9dw3YPlShTBkZDUUG83VrdAXA6ks5rXkXHgaJpZM4KqWGx> > . > — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <https://github.com/waldronlab/MultiAssayExperiment/issues/165#issuecomment-276196089>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEuhn1rnNcafH5uNlvWtBau_W3f6ZfpYks5rXlW9gaJpZM4KqWGx> .

[lgatto]

My 2 cents about what kind of extra features I would be looking for:

• scalability: does the architecture and the manipulation scale to 100s-1000s of samples, each with several assays
• different back-ends: local files, remote files, databases, in-the-cloud data, ... (relates to point above)
• sparse data shouldn't be an issue, or should be easy to filter out
• integration with good annotation: query using a gene name will translate to regions or transcript/protein identifiers according to the type of assay (I remember one of the early vignettes that @vjcitn had but together that really impressed me doing queries along these lines, but I don't remember exactly what it did).
• how easy is it to share the multi-assay data with collaborators, in particular experimentalists (for example through shiny apps)

[aedin]

Hi 1). easy basic filtering, eg remove rows/cols or row/cols with many missing values, or zeros... , easy basic transformation log2/log10 . I think this which and I know there is an easier solution exprs(experiments(ovMAEO)[[1]])<-log2(assay(experiments(ovMAEO)[[1]])) 3) identical, setdiff, intersection a%in% x,y,z across all rows/cols 4) multi-version of summary, give the range min/max/mean/median etc of each experiment, 5. Exploration visualization of MAE data.. shiny finally.. its only as useful as how many people use, it... lets make it as easy as possible for people to use, there are a few place where the help files could have more examples Aedin

…

On 2/2/17 12:57, Laurent Gatto wrote: My 2 cents about what kind of /extra/ features I would be looking for: * scalability: does the architecture and the manipulation scale to 100s-1000s of samples, each with several assays * different back-ends: local files, remote files, databases, /in-the-cloud/ data, ... (relates to point above) * sparse data shouldn't be an issue, or should be easy to filter out * integration with good annotation: query using a gene name will translate to regions or transcript/protein identifiers according to the type of assay (I remember one of the early vignettes that @vjcitn <https://github.com/vjcitn> had but together that really impressed me doing queries along these lines, but I don't remember exactly what it did). * how easy is it to share the multi-assay data with collaborators, in particular experimentalists (for example through shiny apps) — You are receiving this because you commented. Reply to this email directly, view it on GitHub <https://github.com/waldronlab/MultiAssayExperiment/issues/165#issuecomment-277032476>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AArmU-JFiP6ZUmFr-DCMuEjWUywzCW66ks5rYhkHgaJpZM4KqWGx>.

-- Aedin Culhane --------------- Department of Biostatistics and Computational Biology,Dana-Farber Cancer Institute Department of Biostatistics, Harvard TH Chan School of Public Health phone: 617 640 8107 email: aedin@jimmy.harvard.edu

[lwaldron]

Converted to discussion, just so participants know the issue wasn't just closed.