NCBI's Efetch not working

Question

[Jonathan]

Any help would be much appreciated. My goal is to run the following for loop to generate a list of sample_id (which is actually isolation site) for a list of SRAs. However I get an error (see below) for each and every SRA.

 for sra in `awk 'NR>1{print $1}' metadata.txt`
 do
     esearch -db sra -query $sra | efetch -format text | grep "sample_id" | sed 's/^.*G_DNA_//' | sed 's/"//' >> isolation_sites.txt
 done

I also get the same error (see below) when I run it without the for loop, but instead a single line with just a single sample (i.e., no for loop).

 esearch -db sra -query SRS056042 | efetch -format txt

The error is as follows:

curl: (23) Failed writing body

ERROR: curl command failed ( Tue Dec 7 14:17:34 EST 2021 ) with: 23

https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=sra&term=Unable%20%5BACCN%5D%20OR%20to%20%5BACCN%5D%20OR%20locate%20%5BACCN%5D%20OR%20xtract%20%5BACCN%5D%20OR%20executable.%20%5BACCN%5D%20OR%20Please%20%5BACCN%5D%20OR%20execute%20%5BACCN%5D%20OR%20the%20%5BACCN%5D%20OR%20following%20%5BACCN%5D%20OR%20nquire%20%5BACCN%5D%20OR%20dwn%20%5BACCN%5D%20OR%20ftp.ncbi.nlm.nih.gov%20%5BACCN%5D%20OR%20entrez%20%5BACCN%5D%20OR%20entrezdirect%20%5BACCN%5D%20OR%20xtract.Darwin.gz%20%5BACCN%5D%20OR%20gunzip%20%5BACCN%5D%20OR%20f%20%5BACCN%5D%20OR%20xtract.Darwin.gz%20%5BACCN%5D%20OR%20chmod%20%5BACCN%5D%20OR%20x%20%5BACCN%5D%20OR%20xtract.Darwin%20%5BACCN%5D&retmax=10000

HTTP/1.1 200 OK

curl: (22) The requested URL returned error: 400

ERROR: curl command failed ( Tue Dec 7 14:17:35 EST 2021 ) with: 22

https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi -d db=sra&id=x%2Cdwn%2Cf%2Cchmod%2Centrez%2Centrezdirect%2Cftp.ncbi.nlm.nih.gov%2Cgunzip%2Cnquire%2Cxtract.Darwin%2Cxtract.Darwin.gz%2CPlease%2CUnable%2Cexecutable.%2Cexecute%2Cfollowing%2Clocate%2Cthe%2Cto%2Cxtract&rettype=txt&retmode=text&tool=edirect&edirect=16.2&edirect_os=Darwin&email=home%40JJGs-MacBook-Pro.local

HTTP/1.1 400 Bad Request

WARNING: FAILURE ( Tue Dec 7 14:17:34 EST 2021 )

nquire -url https://eutils.ncbi.nlm.nih.gov/entrez/eutils/ efetch.fcgi -db sra -id x,dwn,f,chmod,entrez,entrezdirect,ftp.ncbi.nlm.nih.gov,gunzip,nquire,xtract.Darwin,xtract.Darwin.gz,Please,Unable,executable.,execute,following,locate,the,to,xtract -rettype txt -retmode text -tool edirect -edirect 16.2 -edirect_os Darwin -email [e-mail]

EMPTY RESULT

[Istvan Albert]

The command below

esearch -db sra -query SRS056042 | efetch -format txt

does work for me on MacOS

the error message seems to indicate some sort of installation error.

Unable,executable.,execute,following,locate,the,to,xtrac

Answer 1

[GenoMax]

I don't see that issue on macOS. So you must have a problem with your EntrezDirect install as noted by @Istvan. Best option is to use conda for the install.

$  esearch -db sra -query SRS056042 | efetch -format txt | grep "sample_id" | sed 's/^.*G_DNA_//' | sed 's/"//'
Buccal mucosa of a female participant in the dbGaP study "HMP Core Microbiome Sampling Protocol A (HMP-A)" spots="13744700" bases="2748940000" tax_id="646099" organism="human metagenome"><IDENTIFIERS><PRIMARY_ID>SRS056042</PRIMARY_ID><EXTERNAL_ID namespace="BioSample">SAMN00076051</EXTERNAL_ID><EXTERNAL_ID namespace="dbGaP" label="Sample name">228-700106303</EXTERNAL_ID><EXTERNAL_ID namespace="phs000228" label="submitted sample id">700106303</EXTERNAL_ID></IDENTIFIERS></Member></Pool><CloudFiles><CloudFile filetype="run" provider="gs" location="gs.US"/><CloudFile filetype="run" provider="s3" location="s3.us-east-1"/></CloudFiles><Statistics nreads="2" nspots="13744700"><Read index="0" count="13744700" average="100" stdev="0"/><Read index="1" count="13744700" average="100" stdev="0"/></Statistics><Bases cs_native="false" count="2748940000"><Base value="A" count="32241253"/><Base value="C" count="25742833"/><Base value="G" count="25774267"/><Base value="T" count="32172905"/><Base value="N" count="2633008742"/></Bases></RUN></RUN_SET><ControlledAccess study="phs000228" consent="HMP"/></EXPERIMENT_PACKAGE>

[Jonathan]

Thank you! Looks like it was an installation error, as you suggested. The original install command came from the NCBI website (https://www.ncbi.nlm.nih.gov/books/NBK179288/)

 sh -c "$(curl -fsSL ftp://ftp.ncbi.nlm.nih.gov/entrez/entrezdirect/install-edirect.sh)"

I just re-installed via conda and everything works now as expected.

conda install -c bioconda entrez-direct