search

Lioscro / alaska-parse

Automated and friendly RNA-seq analysis (migrated to Parse Server)
http://alaska.caltech.edu
1 stars 0 forks source link

issue running differential expression analysis on tutorial #27

Open dangeles opened 5 years ago

dangeles commented 5 years ago

Straight from live output. I used the weirdest genotype names I could get. Other than that, should be a standard wt vs mt analysis. It seems that the genotype names may be throwing some errors, but the blocking problem is at the level of target_mapping, which seems really strange. 

# starting sleuth for project ediVSxZNYh
## Rscript /alaska/scripts/sleuth.R -d /alaska/data/projects/ediVSxZNYh/diff -k /alaska/data/projects/ediVSxZNYh/quant -o /alaska/data/projects/ediVSxZNYh/diff -a /alaska/data/organisms/caenorhabditis/elegans/WS268.PRJEB28388/reference/c_elegans_WS268.PRJEB28388_annotation.tsv
# process started 2019-07-26 17:53:29
ediVSxZNYh: Loading required package: methods
ediVSxZNYh: Attaching package: files
ediVSxZNYh: The following object is masked from package:stats:
ediVSxZNYh:     df
ediVSxZNYh: Warning messages:
ediVSxZNYh: 1: In if (j < length(these.flags) & spec[rowmatch, col.has.argument] ==  :
ediVSxZNYh:   the condition has length > 1 and only the first element will be used
ediVSxZNYh: 2: In if (spec[rowmatch, col.has.argument] == flag.no.argument) { :
ediVSxZNYh:   the condition has length > 1 and only the first element will be used
ediVSxZNYh: 3: In if (current.flag == 0) { :
ediVSxZNYh:   the condition has length > 1 and only the first element will be used
ediVSxZNYh: 4: In if (current.flag > 0) { :
ediVSxZNYh:   the condition has length > 1 and only the first element will be used
ediVSxZNYh: 5: In if (substr(peek.optstring, 1, 1) != "-" | (substr(peek.optstring,  :
ediVSxZNYh:   the condition has length > 1 and only the first element will be used
ediVSxZNYh: [1] "#Fetching annotations"
ediVSxZNYh: [1] "# Reading analysis matrix"
ediVSxZNYh: [1] "genotype"
ediVSxZNYh: [[1]]
ediVSxZNYh: [1] "control-wild type"
ediVSxZNYh: [2] "test-hif-1 Ex[P_myo2::GFP,P_unc22::RFP]"
ediVSxZNYh: [1] "# 1 conditions detected"
ediVSxZNYh: [1] "# Appending path to kallisto result folders."
ediVSxZNYh:   sample                                genotype
ediVSxZNYh: 1    wt2                       control-wild type
ediVSxZNYh: 2    mt1 test-hif-1 Ex[P_myo2::GFP,P_unc22::RFP]
ediVSxZNYh: 3    wt1                       control-wild type
ediVSxZNYh: 4    mt2 test-hif-1 Ex[P_myo2::GFP,P_unc22::RFP]
ediVSxZNYh:                                         path
ediVSxZNYh: 1 /alaska/data/projects/ediVSxZNYh/quant/wt2
ediVSxZNYh: 2 /alaska/data/projects/ediVSxZNYh/quant/mt1
ediVSxZNYh: 3 /alaska/data/projects/ediVSxZNYh/quant/wt1
ediVSxZNYh: 4 /alaska/data/projects/ediVSxZNYh/quant/mt2
ediVSxZNYh: [1] "# Creating Sleuth object."
ediVSxZNYh: reading in kallisto results
ediVSxZNYh: dropping unused factor levels
ediVSxZNYh: ....
ediVSxZNYh: Error in check_target_mapping(tmp_names, target_mapping, !is.null(aggregation_column)) :
ediVSxZNYh:   couldn't solve nonzero intersection
ediVSxZNYh: Calls: sleuth_prep -> check_target_mapping
ediVSxZNYh: Execution halted
# process finished 2019-07-26 17:53:32
command terminated with non-zero return code!
dangeles commented 5 years ago

it appears this problem has been referenced by other users and is the result of inadequate mapping between the target_ids and other transcript identifiers.

Lioscro commented 5 years ago

Is this an issue with sleuth?

dangeles commented 5 years ago

it's probably an issue with the way the genomics files are being formatted... but i'm not sure

Lioscro commented 5 years ago

I have found out why. It's because the target_ids in the reference CDNA don't match with that in the annotation.

The CDNA fasta obtained from Wormbase for PRJEB28388 has target_ids and genes of this form >chrIII_pilon.g6684 gene=chrIII_pilon.g6684, but the annotation tsv we received from Michael has target_ids of this form 2L52.1a.1.

I'll ask why they don't match and if they could give us annotations with matching target_ids.

Lioscro commented 5 years ago

For the time being, I think it will work if you change the organism to WS268.PRJNA13758 and rerun Alaska.

dangeles commented 5 years ago

Ok. Let's make only the working genome available until fixed.

Sent from my iPhone

On Aug 18, 2019, at 9:51 AM, Joseph Min notifications@github.com<mailto:notifications@github.com> wrote:

For the time being, I think it will work if you change the organism to WS268.PRJNA13758 and rerun Alaska.

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHubhttps://github.com/Lioscro/alaska-parse/issues/27?email_source=notifications&email_token=ABRVIETQJ3CBORYBZY3T6TTQFFHU7A5CNFSM4IHGAJTKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD4RAP2Y#issuecomment-522323947, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ABRVIESGUOI3ETPE5YTUICDQFFHU7ANCNFSM4IHGAJTA.

dangeles commented 5 years ago

And let's make sure it's fixed for all species

Sent from my iPhone

On Aug 18, 2019, at 9:51 AM, Joseph Min notifications@github.com<mailto:notifications@github.com> wrote:

For the time being, I think it will work if you change the organism to WS268.PRJNA13758 and rerun Alaska.

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHubhttps://github.com/Lioscro/alaska-parse/issues/27?email_source=notifications&email_token=ABRVIETQJ3CBORYBZY3T6TTQFFHU7A5CNFSM4IHGAJTKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD4RAP2Y#issuecomment-522323947, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ABRVIESGUOI3ETPE5YTUICDQFFHU7ANCNFSM4IHGAJTA.

Lioscro commented 5 years ago

Should be able to fix this now with #49