error when searching genename in epgg/eg docker image

[jos4uke]

Hi

I tried to search genename on hg19 genome but the server crashed as reported in the log window.

# /var/log/apache2/error.log
subtileKnife starting... >>>>>
cgi param parsed
Hooked to hg19
reviveScaffold() done
subtleKnife: subtleKnife.c:5164: queryGene: Assertion `fin!=NULL' failed.

Could you tell me how to index correctly genename?

[jos4uke]

Hi

I took some time to inspect the error and I traced the calls from the debugger. When searching a gene name "NAC001" , I caught the server response telling there is a json syntax error:

# header
URL= http://localhost/cgi-bin/subtleKnife?getcoord4genenames=on&scaffoldruntimenoupdate=on&dbName=tair10&lst=NAC&searchgenetknames=tair10Gene&session=pBNDrKbGOH&statusId=1&hmspan=820

# server response
GETXHRhttp://localhost/cgi-bin/subtleKnife[HTTP/1.1 200 OK 64ms] 
Réponse
SyntaxError: JSON.parse: expected property name or '}' at line 1 column 2 of the JSON data
Charge utile de la réponse
{result:[],}

I checked for the gene name "NAC001" in mysql, dbName=tair10 in table tair10Gene, the gene name exists:

mysql> select * from tair10Gene where name = "NAC001";
+-------+-------+------+--------+
| chrom | start | stop | name   |
+-------+-------+------+--------+
| Chr1  |  3631 | 5899 | NAC001 |
+-------+-------+------+--------+
1 row in set (0.00 sec)

I am not sure why the ajax request did not return the gene name present in the database? Your help is welcome.

[lidaof]

Hi @jos4uke , sorry I forgot to reply that day...

I think the issue is caused by missing gene.gz and .tbi file on the /srv/epgg folder.

say refGene.gz and .tbi need be located in /srv/epgg/data/data/subtleKnife/hg19

[jos4uke]

Hi @lidaof

I checked for the files you mentionned, and they are present:

# inside the container
root@30e354f9d603:/srv/epgg/data/data/subtleKnife/hg19# ls
config  gencodeV17.gz  gencodeV17.gz.tbi  refGene.gz  refGene.gz.tbi  session  xenoRefGene.gz  xenoRefGene.gz.tbi

but searching for gene name "HOXA1" for instance returns "No result" while the log noticed me "Server crashed". The gene is present in refGene table as shown below.

mysql> select * from refGene where name like "HOXA1";
+-------+----------+----------+-------+
| chrom | start    | stop     | name  |
+-------+----------+----------+-------+
| chr7  | 27132613 | 27135625 | HOXA1 |
| chr7  | 27132613 | 27135625 | HOXA1 |
+-------+----------+----------+-------+
2 rows in set (0.00 sec)

My previous example with "NAC001" was done on arabidopsis and the same error occured. And reference genes are in tair10Gene.gz and .tbi files.

# inside the container
root@30e354f9d603:/srv/epgg/data/data/subtleKnife/tair10# ls
config  gc5Base.bigWig  repeat.gz  repeat.gz.tbi  scaffoldInfo  session  tair10Gene.gz  tair10Gene.gz.tbi  tair10.gz  tair10.gz.tbi

[jos4uke]

@lidaof Do you think the trailing comma in the json object {result:[],} caused the error? Well it didn't explain why the result is null whereas the gene name exists.

[lidaof]

oh, if it's the docker image the files should be there. While I don't think the reason was by trailing comma...are you using the docker from here https://hub.docker.com/r/epgg/eg/ ?

[jos4uke]

yes absolutely hg19 files were already there, and I added tair10 for evaluation

[lidaof]

my bad...

seems the gencodeV19 files didn't exist caused this problem

in /srv/epgg/data/data/subtleKnife/hg19/config/tracks.json, please change dbsearch to false like:

        {name:'gencodeV19',label:'GENCODE V19 genes',
                filetype:'hammock',
                categories:{1:['coding','rgb(0,60,179)'], 2:['non-coding','rgb(0,128,0)'], 3:['pseudogene','rgb(230,0,172)'], 4:['problem','rgb(255,0,0)'], 5:['polyA','rgb(0,0,51)']},
                url:'http://egg.wustl.edu/d/hg19/gencodeV19.gz',
                isgene:true,dbsearch:false },
        {name:'gencodeV17',label:'GENCODE V17 genes',
                filetype:'hammock',
                categories:{1:['coding','rgb(0,60,179)'], 2:['non-coding','rgb(0,128,0)'], 3:['pseudogene','rgb(230,0,172)'], 4:['problem','rgb(255,0,0)'], 5:['polyA','rgb(0,0,51)']},
                url:'http://egg.wustl.edu/d/hg19/gencodeV17.gz',
                isgene:true,dbsearch:false },

[jos4uke]

Thanks a lot

it solved the error for hg19

but do you have an idea why it does not work for tair10?

Here is my config:

# /srv/epgg/data/data/subtleKnife/tair10/config/tracks.json
{

Genes : [
        {name:'tair10Gene',label:'TAIR genes',
                filetype:'hammock',
                url:'http://egg.wustl.edu/d/tair10/tair10Gene.gz',
                queryUrl:'http://www.arabidopsis.org/servlets/TairObject?type=gene&name=',
                boxcolor:'#002EB8',
                isgene:true,dbsearch:true },
],

Repeat : [
        {name:'repeat',label:'Transposable elements',
                filetype:'bed',
                url:'http://egg.wustl.edu/d/tair10/repeat.gz',
        },
],

'G/C related' : [
        {name:'gc5Base',label:'GC percent',
                filetype:'bigwig',
                url:'http://egg.wustl.edu/d/tair10/gc5Base.bigWig',
                height:50,
                colorpositive:'#4785C2/#4747C2',
                fixedscale:{min:0,max:100},
        },
],

}

[lidaof]

you do have tair10Gene in mysql, and tair10Gene.gz and tair10Gene.tbi in the epgg/tair10 folder, right? What's the error log from Apache when you search on arabidopsis?

[jos4uke]

Yes I have all tair10 files in the container

# inside the container
root@30e354f9d603:/srv/epgg/data/data/subtleKnife/tair10# ls
config  gc5Base.bigWig  repeat.gz  repeat.gz.tbi  scaffoldInfo  session  tair10Gene.gz  tair10Gene.gz.tbi  tair10.gz  tair10.gz.tbi

# search ECS1 in tair10Gene.gz
root@30e354f9d603:/srv/epgg/data/data/subtleKnife/tair10# zgrep "ECS1" tair10Gene.gz
Chr1    11311169    11311989    name:"AT1G31580.1",id:3186,strand:"+",struct:{thin:[[11311169,11311300],[11311712,11311989],],thick:[[11311300,11311405],[11311531,11311712],],},desc:"",name2:"ECS1"

# search ECS1 in mysql
mysql> select * from tair10Gene where name = "ECS1";
+-------+----------+----------+------+
| chrom | start    | stop     | name |
+-------+----------+----------+------+
| Chr1  | 11311169 | 11311989 | ECS1 |
+-------+----------+----------+------+
1 row in set (0.01 sec)

Here is the log error from apache2 when I searched for "ECS1" tair10 gene

# /var/log/apache2/error.log 
done
subtileKnife starting... >>>>>
cgi param parsed
Hooked to tair10
done
subtileKnife starting... >>>>>
cgi param parsed
Hooked to tair10
reviveScaffold() done
done
subtileKnife starting... >>>>>
cgi param parsed
Hooked to tair10
done
subtileKnife starting... >>>>>
cgi param parsed
Hooked to tair10
done
subtileKnife starting... >>>>>
cgi param parsed
Hooked to tair10
done
subtileKnife starting... >>>>>
cgi param parsed
Hooked to tair10
done
subtileKnife starting... >>>>>
cgi param parsed
Hooked to tair10
reviveScaffold() done
done

Seems no error occured, and the log window in epigenome browser said "No genes found" as shown in this screenshot:

I also checked in mysql log error, and here are the corresponding lines:

# 2018-03-29T07:51:46.930141Z 213 [Note] Aborted connection 213 to db: 'tair10' user: 'root' host: 'localhost' (Got an error reading communication packets)
2018-03-29T07:51:47.207258Z 214 [Note] Aborted connection 214 to db: 'tair10' user: 'root' host: 'localhost' (Got an error reading communication packets)
2018-03-29T07:51:49.230311Z 215 [Note] Aborted connection 215 to db: 'tair10' user: 'root' host: 'localhost' (Got an error reading communication packets)
2018-03-29T07:51:49.633480Z 216 [Note] Aborted connection 216 to db: 'tair10' user: 'root' host: 'localhost' (Got an error reading communication packets)
2018-03-29T07:51:57.597037Z 217 [Note] Aborted connection 217 to db: 'tair10' user: 'root' host: 'localhost' (Got an error reading communication packets)

But I am not sure this is related to this issue, because when I searched by coordinates the same kind of lines appear in the log but the browser jumps to it.

[jos4uke]

When searching for "HOXA1" hg19 gene, I got the same log lines:

# /var/log/apache2/error.log
subtileKnife starting... >>>>>
cgi param parsed
Hooked to hg19
reviveScaffold() done
done
subtileKnife starting... >>>>>
cgi param parsed
Hooked to hg19
reviveScaffold() done
done

and in mysql

# /var/log/mysql/error.log
2018-03-29T07:56:41.637292Z 272 [Note] Aborted connection 272 to db: 'hg19' user: 'root' host: 'localhost' (Got an error reading communication packets)
2018-03-29T07:56:45.912621Z 273 [Note] Aborted connection 273 to db: 'hg19' user: 'root' host: 'localhost' (Got an error reading communication packets)

and the browser returned correctly a list of all HOXA1 pattern matching genes, as shown here:

[lidaof]

hi @jos4uke, currently I don't know the reason now..would you mind restart mysql see it it's working? or re-dump the gene table on tair10 and restart mysql? Thanks....

[jos4uke]

hi @lidaof

I tried as you recommended to restart and rebuild tair10 but

• restart mysql then search gene name failed
• dump tair10 database, rebuild tair10, then search for gene name failed

I also restarted apache but search gene name did not work

I don't know what to try to make it work .... :(

[lidaof]

Hi @jos4uke , ok, sorry for that. Would you mind pull the latest epgg docker from docker hub and re-run the tail10 genome? I just updated the docker image yesterday. Thanks a lot.

[jos4uke]

Hi @lidaof

You are the boss :+1: :)

[jos4uke]

When I searched for ECS1 gene in tair10, it works and I only got a message in the log window, for information:

atbplevel but totallen>hmSpan*3

[lidaof]

Hi @jos4uke , that log seems not relevant to this issue, I am not aware of other reasons now...sorry...