Error: No such file or directory: python2.7/site-packages/altanalyze/Config//arrays.txt

[therealgenna]

I just installed AltAnalyze through pip in conda environment (with python 2.7) and following instructions here tried to install the organism database with
altanalyze --species Dr --update Official --additional all --version EnsMart102

This threw the following error (I've replaced some paths with /path/to/):

Current database version: EnsMart102
AltAnalyze.py --species Dr --update Official --additional all --version EnsMart102
Traceback (most recent call last):
  File "AltAnalyze.py", line 8435, in runCommandLineVersion
    commandLineRun()
  File "AltAnalyze.py", line 7951, in commandLineRun
    array_codes = UI.remoteArrayInfo()
  File "/path/to/lib/python2.7/site-packages/altanalyze/UI.py", line 3679, in remoteArrayInfo
    importArrayInfo()
  File "/path/to/lib/python2.7/site-packages/altanalyze/UI.py", line 3685, in importArrayInfo
    for line in open(fn,'rU').readlines():
IOError: [Errno 2] No such file or directory: '/path/to/lib/python2.7/site-packages/altanalyze/Config//arrays.txt'

I don't know what should I do with it. Any help is appreciated.

[therealgenna]

Before the above, with --version EnsMart102, I tried without the --version : altanalyze_ICGS2]$ altanalyze --species Dr --update Official --additional all and got this error:

AltAnalyze depedency not met for: wx
...The AltAnalyze Results Viewer requires wx

WARNING!!!! Some dependencies are not currently met.
This may impact AltAnalyze's performance

Running commandline options

Arguments input: ['AltAnalyze.py', '--species', 'Dr', '--update', 'Official', '--additional', 'all'] 

AltAnalyze.py --species Dr --update Official --additional all
Downloading the following file: species_all.txt   Connection opened. Downloading (be patient)
* 

Downloading the following file: source_data.txt   Connection opened. Downloading (be patient)
* 

Downloading the following file: array_versions.txt   Connection opened. Downloading (be patient)
* * 

Finished downloading the latest configuration files.
Species name to update: Danio rerio
Ensembl version current
Traceback (most recent call last):
  File "AltAnalyze.py", line 8435, in runCommandLineVersion
    commandLineRun()
  File "AltAnalyze.py", line 7975, in commandLineRun
    for ad in db_versions_vendors[db_version]:
KeyError: 'current'

[nsalomonis]

EnsMart102 is not currently provided unless you end up building it on your end which is possible. Can you run with --version EnsMart72 ? This is an older version of Ensembl but the default supported version (EnsMart100 is build an in testing but not yet uploaded). If required, let me know. Best, Nathan

…

On Mon, Dec 14, 2020 at 2:52 PM therealgenna ***@***.***> wrote: Before the above, with --version EnsMart102, I tried without the --version : altanalyze_ICGS2]$ altanalyze --species Dr --update Official --additional all and got this error: AltAnalyze depedency not met for: wx ...The AltAnalyze Results Viewer requires wx WARNING!!!! Some dependencies are not currently met. This may impact AltAnalyze's performance Running commandline options Arguments input: ['AltAnalyze.py', '--species', 'Dr', '--update', 'Official', '--additional', 'all'] AltAnalyze.py --species Dr --update Official --additional all Downloading the following file: species_all.txt Connection opened. Downloading (be patient) * Downloading the following file: source_data.txt Connection opened. Downloading (be patient) * Downloading the following file: array_versions.txt Connection opened. Downloading (be patient) * * Finished downloading the latest configuration files. Species name to update: Danio rerio Ensembl version current Traceback (most recent call last): File "AltAnalyze.py", line 8435, in runCommandLineVersion commandLineRun() File "AltAnalyze.py", line 7975, in commandLineRun for ad in db_versions_vendors[db_version]: KeyError: 'current' — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#48 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABD5TSKBWJL4QHBFERTYY6LSUZUJLANCNFSM4U3HKBOQ> .

[therealgenna]

Now I see that arrays.txt sits in a different place, under my home directory ~/altanalyze/Config/, and this is not where the script is looking for it - in /path/to/lib/python2.7/site-packages/altanalyze/Config//arrays.txt which is under the python installation! Under pythin installation, in the Config dir I have the following files:
array_versions.txt report.log source_data.txt species_all.txt version.txt

[therealgenna]

Hi Nathan, thank you for a quick response! No, just tried - it's the same with --version EnsMart72. I think it's because different files got into different places, as I posted above. Where should the files like arrays.txt be? It seems strange that the program is looking for it under the python directory tree ...

[nsalomonis]

The files should be OK in either place, but it’s possible we are missing something. As a sanity check, can you download the software from GitHub from https://github.com/nsalomonis/altanalyze and try again?

…

On Mon, Dec 14, 2020 at 4:42 PM therealgenna ***@***.***> wrote: Hi Nathan, thank you for a quick response! No, just tried - it's the same with --version EnsMart72. I think it's because different files got into different places, as I posted above. Where should the files like array.txt be? It seems strange that the program is looking for it under the python directory tree ... — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#48 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABD5TSMAJSU7N4DMSCQVGVDSU2BCVANCNFSM4U3HKBOQ> .

[therealgenna]

I downloaded, unzipped and ran python AltAnalyze.py --species Dr --update Official --additional all --version EnsMart72 which worked. Is EnsMart72 the latest I can use? I don't want to burden you with creating a newer version if that's the latest - I am not sure I'll need it. I just want to run ICGS2 on scRNAseq to do clustering.

So would you recommend me to just stick with the downloaded version and ignore the one with pip install altanalyze?

[nsalomonis]

It shouldn’t honestly matter too much (better to have the latest GitHub version honestly), since it will primarily impact default pathway analyses. Are you combining multiple h5, mtx or text files? If so, there is a simple option to do this by adding all of the input files to one folder and setting that as the target path for ChromiumSparseMatrix Having R and Biobase and hopach in R will improve the clustering. Best, Nathan

…

On Mon, Dec 14, 2020 at 5:12 PM therealgenna ***@***.***> wrote: I downloaded, unzipped and ran python AltAnalyze.py --species Dr --update Official --additional all --version EnsMart72 which worked. Is EnsMart72 the latest I can use? I don't want to burden you with creating a newer version if that's the latest - I am not sure I'll need it. I just want to run ICGS2 on scRNAseq to do clustering. So would you recommend me to just stick with the downloaded version and ignore the one with pip install altanalyze? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#48 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABD5TSNCVEIUNA6Z3FV5P7LSU2ETDANCNFSM4U3HKBOQ> .

[therealgenna]

Great, thank you so much! I guess I'll have to extract count matrices and make text files from Seurat objects which were made/merged from multiple 10X run outputs.

[therealgenna]

FYI, I noticed that with pip install I get Config.zip and AltDatabase.zip in the /path/to/python2.7/site-packages/altanalyze but there are no unzipped versions [update: there is a Config dir here, but it only contains version.txt file]. When I try to install the database as above, the ~/altanalyze is created in my home directory, under which I also get ~/altanalyze/Config and ~/altanalyze/AltDatabase, but then I get the reported error.

Moving Config and AltDatabase from ~/altanalyze to /path/topython2.7/site-packages/altanalyze works (and I guess unzipping the zip files would work as well).

[nsalomonis]

That is fine or you can add all Cell Ranger sparse matrix outputs to one folder with the appropriate sample names added (e.g., cancer_rep1.h5). ICGS will merge all them appropriately. If you use a counts tab-delimited text file from Seurat as input, include the option:

…

--dataFormat counts If more than 20k cells I suggest:

--downsample 5000 If the data is already scaled and log2, you can exclude the counts command. When the data is in sparse matrix format it knows to consider it counts. Best, Nathan

On Dec 14, 2020, at 6:14 PM, therealgenna ***@***.***> wrote:  Great, thank you so much! I guess I'll have to extract count matrices and make text files from Seurat objects which were made/merged from multiple 10X run outputs. — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

[nsalomonis]

I would still recommend using the GitHub version which addresses a few bugs present in the last pypi release. Much more stable. The zip files you mentioned are correct. The site-packages path can’t be modified so those folders are extracted to your root user directory and should be looked at by the program first when running which is the expected behavior for the PyPi. Not sure what the issue is yet. Best, Nathan

…

On Dec 14, 2020, at 9:04 PM, therealgenna ***@***.***> wrote:  FYI, I noticed that with pip install I get Config.zip and AltDatabase.zip in the /path/topython2.7/site-packages/altanalyze but there are no unzipped versions. When I try to install the database as above, the ~/altanalyze is created in my home directory, under which I also get ~/altanalyze/Config and ~/altanalyze/AltDatabase, but then I get the reported error. Moving Config and AltDatabase from ~/altanalyze to /path/topython2.7/site-packages/altanalyze works (and I guess unzipping the zip files would work as well). — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

[therealgenna]

Awesome, I will follow your recommendations. In the end, after testing it on smaller sets, I might run it on ~400k cells. Will it work? What number should I use for downsampling?

Another question - I've emailed you a day or two ago asking about Louvain resolution in your ICGS2 paper. I wonder if you had a chance to see my email.

Thanks again!

[nsalomonis]

Sorry for missing this before. Does the below help (ICGS_NMF.py script):

from annoy import AnnoyIndex
t=AnnoyIndex(nm,metric="euclidean")
for i in range(nn):
    try:  t.add_item(i,X[i])
    except Exception: print i
t.build(100)
diclst={}
for i in range(nn):
    ind=t.get_nns_by_item(i,10)
    diclst[i]=ind
G=nx.from_dict_of_lists(diclst)
dendrogram= community.generate_dendrogram(G)
level=0
pr= community.partition_at_level(dendrogram,level)