September 2007 - Mg-rast - mailman3.cels.anl.gov

Re: mg-rast Query
by Daniel Paarmann 28 Sep '07

28 Sep '07

Hi Jack, > (2) Also, the analysis of jobs 207 and 211 seems to of suddenly > changed with the latest addition of sequences. Preliminary analysis > of the first 50% of the reads produced around 50% of the sequences > identified in the subsystems, when I added the remaining 50% of > sequences from the run, the analysis showed only 11-15% of the > sequences hitting subsystems. I hope this is an error. > I compared the two new jobs with the previously computed results. 4440164.4 - Bag1_19thMay_cDNA (Sep 17 2007) Number of sequences uploaded: 1509 from the overview page: Number of Fragments 169 Number of Subsystems 31 in Subystems: 14% 4440164.5 - Bag1_19thMay_cDNA (Sep 20 2007) Number of sequences uploaded: 35712 from the overview page: Number of Fragments 6384 Number of Subsystems 297 in Subsystems: 11% 4440166.3 - Bag6_19thMay_cDNA (Sep 13 2007) Number of sequences uploaded: 12940 from the overview page: Number of Fragments 2057 Number of Subsystems 187 in Subsystems: 15% 4440166.5 - Bag6_19thMay_cDNA (Sep 20 2007) Number of sequences uploaded: 99488 from the overview page: Number of Fragments 18315 Number of Subsystems 407 in Subsystems: 15% Overall, the ability of the metagenomics RAST server to associate those fragments with subsystems and functional roles has not changed. I suspect that the issue of browsing multiple RAST jobs at the same time in the SeedViewer has caused this. Both Bag1_13thMay_cDNA (Sep 17 2007) and Bag6_13thMay_cDNA (Sep 17 2007) had a very high number of features in subsystems (55% and 49% respectively). If you were comparing all the initials samples in multiple browser windows (or tabs), the displayed data would only match the latest RAST job. As I tried to explain in my previous mail, this is a known and critical issue and will be addressed with the upcoming release of the SeedViewer 2.0. - Daniel

1 0

mg-rast Re: Questions about mg rast
by Daniel Paarmann 26 Sep '07

26 Sep '07

Hi Jack, > i have figured out the problem with the identical taxanomic > classifications. It seems that you cannot open two different > organism overviews without the second one imposing it's tax > classification on the first. > This is not a browser issue, but caused by the way how the RAST Server exchanges information with the SeedViewer. Since this technology is inherited from the regular RAST server which focuses on annotations of single organisms, the comparison of two (metagenome) RAST jobs _at the same time_ is not yet sufficiently covered. As soon as you open a SeedViewer from the mg-RAST Job Details page, this SeedViewer provides you the views for this metagenome. Unfortunately when you open another SeedViewer from a different Job Details page, your SeedViewer pages will display the data from that metagenome from there on. This is a known problem we are about to address with the upcoming release of the SeedViewer 2.0. > very odd...but appears to be a brower issue? I use firefox as many > of the gifs wonn't open in IE. > We did some testing for Internet Explorer and have fixed the problem with graphics not loading in the new release. For now the best browser to use is firefox.

1 0

mg-rast Re: Questions about mg rast
by Daniel Paarmann 26 Sep '07

26 Sep '07

Jack, > The two other samples have still not completed...is there something > holding them up??? Although we re-started job 206 on friday, it failed again in the similarity post processing step. Our preliminary analysis of the error logs shows a problem new to us and we need more time to look into it. Job 212 is running on the cluster. The cluster is currently very busy and jobs will be computed in the order they come in, prioritizing regular RAST jobs over metagenome RAST jobs. Daniel

1 0

mg-rast Query
by Jack Gilbert 26 Sep '07

26 Sep '07

Dear MG-RAST techs, I have a couple of problems with the SEED database analysis interface. (1) It appears that from the SEED homepage I can run a BLAST search on my data if I approach it via the organism overview, e.g.if I am on a SEED VIEWER Organism overview page for example for run 193. Then when I click on THE SEED link (top right hand corner next to "you are here", it brings up the SEED Viewer start page (http://metagenomics.theseed.org/index.cgi) If I click on BLAST and then select an organism, the Metagenome RAST run (193) appears with it's name "Bag 6 - 632" however, it will not run a BLAST action on it. Is this supposed to happen, or are you not able to BLAST currently. (2) Also, the analysis of jobs 207 and 211 seems to of suddenly changed with the latest addition of sequences. Preliminary analysis of the first 50% of the reads produced around 50% of the sequences identified in the subsystems, when I added the remaining 50% of sequences from the run, the analysis showed only 11-15% of the sequences hitting subsystems. I hope this is an error. (3) A number of the new jobs I have uploaded have failed to complete or are still being analysed. As these were both added last Thursday it seems odd that they would take this long, please can you investigate. Cheers Jack Gilbert (part of Dawn Fields group) -- Dr. Jack A. Gilbert B.Sc. Ph.D. Molecular Ecologist Plymouth Marine Laboratory, Prospect Place, The Hoe, Plymouth, PL1 3DH, UK. Work Tel: +44 (0)1752 633 416 Home Tel: +44 (0)1752 255 783 Mobile Tel: +44 (0)7764 190 147 Email: gilbertjack(a)gmail.com

1 0

mg-rast metagenomics RAST - your jobs
by Daniel Paarmann 24 Sep '07

24 Sep '07

Hello Artur Silva, we are sorry that the metagenomics RAST server had problems processing your job(s). We are still investigating the cause for this error. Unrelated to this software problem in the mg-RAST preprocessing, we found that your tar archive contained 61 FASTA sequence files with one sequence each. The metagenomics RAST server interprets that as 61 metagenome samples and not as a single project consisting of 61 sequences. Consequently, it created lots of jobs from your single tar archive. Given only one sequence the job result would be very limited. I suspect that this was not intended on your side. To upload a set of sequences as a single job, you would have to put them into one FASTA file, which may optionally be in a compressed format (tar/gzip). I've done this on your behalf and removed the old jobs 213-274. Job No. 279 has your 61 sequences from the uploaded tar archive and is currently being computed. If you have any further questions about uploading data to the metagenomics RAST server, please do contact us at mg-rast(a)mcs.anl.gov. Again, our apologies for any inconveniences that may have caused, best regards, Daniel Paarmann

1 0

mg-rast Re: Questions about mg rast
by Jack Gilbert 22 Sep '07

22 Sep '07

Dan, The Blast problem occurs when I do this. If I am on a SEED VIEWER Organism overview page for example for run 193. Then when I click on THE SEED link (top right hand corner next to "you are here", it brings up the SEED Viewer start page (http://metagenomics.theseed.org/index.cgi) If I click on BLAST and then select an organism, the Metagenome RAST run (193) appears with it's name "Bag 6 - 632" however, as I said before it will not run a BLAST action on it. I presume based on your last email that the fact that I can see this is an error? A couple of more questions: Firstly, The analysis of 207 and 211 seems to of suddenly changed with the latest addition of sequences. We now have only 11-15% of the sequences hitting subsystems...whereas with the preliminary analysis of half the data we had some 50% of the sequences hitting subsystems....this is rather odd. Secondly, the problem I refer to is this. The analysis of taxanomic classification of 207 ( http://metagenomics.theseed.org/index.cgi?action=ShowMetaGenomeStatistics&g…) appears to be identical to that of 211 ( http://metagenomics.theseed.org/index.cgi?action=ShowMetaGenomeStatistics&g…) As I said previously this also seem to apply to 193 ( http://metagenomics.theseed.org/index.cgi?action=ShowMetaGenomeStatistics&g…) and 194 ( http://metagenomics.theseed.org/index.cgi?action=ShowMetaGenomeStatistics&g… ) Cheers Jack On 9/21/07, Daniel Paarmann <paarmann(a)mcs.anl.gov> wrote: > > Hello Jack, > > > Virtually all of them would be good to delete except two. > > Therefore, please could you remove: > > > > 183, 184, 186, 187, 188, 189, 190, 191, 192, 195, 198, 199, 200, 201 > > all your those preliminary runs have been removed. > > > One more thing the BLAST search tool for SEED, when I search for > > things that are apparently there, the BLAST search results show > > nothing...it doesn't say anything at all...like it has > > failed....can you comment? > > Unfortunately I cannot reproduce that error. The Seed Viewer start > page (http://metagenomics.nmpdr.org/index.cgi) provides a BLAST > search against current SEED organisms, that however does not include > any sequence processed by the metagenome RAST server. > If you could please provide more detailed information how you've > triggered that problem, I will look into it again. > > > Your job no. 206 had a problem during the post processing of the > similarity computations. It has been restarted and hopefully will run > through this time. We will keep monitoring it. > > > Best regards, > Daniel Paarmann > > > > > > Jack > > > > > > On 9/20/07, Jack Gilbert <gilbertjack(a)gmail.com> wrote: > > Dan, > > > > I have another one with the same problem, however, the compressed > > file is still 14Mg. > > > > Let me know what to do. > > > > Jack > > > > > > On 9/20/07, Jack Gilbert <gilbertjack(a)gmail.com > wrote: > > Dan, > > > > Have also been having problems uploading this file (I have attached > > it compressed). It says each time that it cannot detect FASTA in > > the file. Obviously they are FASTA..so whats going on? > > > > Jack > > > > > > On 9/20/07, Jack Gilbert <gilbertjack(a)gmail.com > wrote: > > Dan, > > > > Virtually all of them would be good to delete except two. > > Therefore, please could you remove: > > > > 183, 184, 186, 187, 188, 189, 190, 191, 192, 195, 198, 199, 200, 201 > > > > These were all preliminary runs and they are getting things > > confused now. will upload the new ones now. > > > > Cheers > > > > Jack > > > > On 9/19/07, Daniel Paarmann <paarmann(a)mcs.anl.gov > wrote: > > Jack, > > > > > the submission of sequences to SEED appears to of changed. It is > > > coming up with a genome centric submission which seems to have no > > > place for metagenomic samples. > > > > thank you for notifying us about this. It has been corrected and the > > metagenome RAST server shows the correct upload page again. > > > > Unfortunately, due to this configuration error on our side, your jobs > > 202-205 have been created as regular RAST jobs. I had to delete them > > and would ask you to upload them again. My apologies for any > > inconvenience this may cause. > > 202 All All All > > 203 Bag1 13thMay cDNA > > 204 Bag6 19thMay cDNA > > 205 Bag1 19thMay cDNA > > > > > Also, quick question regarding the taxanomic analysis, the > > > taxanomic analysis data keeps on changing - at least everytime I > > > look at it I see problems or things that have changed. Is this in > > > continuous flux? > > > > The taxonomic analysis and the databases used for it have not been > > changed since the server release. We appreciate any feedback about > > inconsistencies and will investigate it immediately. Could you be > > more specific for what metagenome (job) you noticed differences? What > > kind of changes did you observe? > > > > Regarding your question how to delete jobs, this action is currently > > not supported by the RAST web service, but I will suggest it for our > > next release. For now, if you send my the job numbers and metagenome > > ids, I'll gladly delete them for you. > > > > > > Best regards, > > Daniel Paarmann > > > > > > > > -- > > Dr. Jack A. Gilbert B.Sc. Ph.D. > > Molecular Ecologist > > Plymouth Marine Laboratory, > > Prospect Place, The Hoe, > > Plymouth, PL1 3DH, UK. > > Work Tel: +44 (0)1752 633 416 > > Home Tel: +44 (0)1752 255 783 > > Mobile Tel: +44 (0)7764 190 147 > > Email: gilbertjack(a)gmail.com > > > > > > > > -- > > Dr. Jack A. Gilbert B.Sc. Ph.D. > > Molecular Ecologist > > Plymouth Marine Laboratory, > > Prospect Place, The Hoe, > > Plymouth, PL1 3DH, UK. > > Work Tel: +44 (0)1752 633 416 > > Home Tel: +44 (0)1752 255 783 > > Mobile Tel: +44 (0)7764 190 147 > > Email: gilbertjack(a)gmail.com > > > > > > > > > > -- > > Dr. Jack A. Gilbert B.Sc. Ph.D. > > Molecular Ecologist > > Plymouth Marine Laboratory, > > Prospect Place, The Hoe, > > Plymouth, PL1 3DH, UK. > > Work Tel: +44 (0)1752 633 416 > > Home Tel: +44 (0)1752 255 783 > > Mobile Tel: +44 (0)7764 190 147 > > Email: gilbertjack(a)gmail.com > > > > > > > > -- > > Dr. Jack A. Gilbert B.Sc. Ph.D. > > Molecular Ecologist > > Plymouth Marine Laboratory, > > Prospect Place, The Hoe, > > Plymouth, PL1 3DH, UK. > > Work Tel: +44 (0)1752 633 416 > > Home Tel: +44 (0)1752 255 783 > > Mobile Tel: +44 (0)7764 190 147 > > Email: gilbertjack(a)gmail.com > > -- Dr. Jack A. Gilbert B.Sc. Ph.D. Molecular Ecologist Plymouth Marine Laboratory, Prospect Place, The Hoe, Plymouth, PL1 3DH, UK. Work Tel: +44 (0)1752 633 416 Home Tel: +44 (0)1752 255 783 Mobile Tel: +44 (0)7764 190 147 Email: gilbertjack(a)gmail.com

1 0

mg-rast Re: Questions about mg rast
by Daniel Paarmann 21 Sep '07

21 Sep '07

Hello Jack, > Virtually all of them would be good to delete except two. > Therefore, please could you remove: > > 183, 184, 186, 187, 188, 189, 190, 191, 192, 195, 198, 199, 200, 201 all your those preliminary runs have been removed. > One more thing the BLAST search tool for SEED, when I search for > things that are apparently there, the BLAST search results show > nothing...it doesn't say anything at all...like it has > failed....can you comment? Unfortunately I cannot reproduce that error. The Seed Viewer start page (http://metagenomics.nmpdr.org/index.cgi) provides a BLAST search against current SEED organisms, that however does not include any sequence processed by the metagenome RAST server. If you could please provide more detailed information how you've triggered that problem, I will look into it again. Your job no. 206 had a problem during the post processing of the similarity computations. It has been restarted and hopefully will run through this time. We will keep monitoring it. Best regards, Daniel Paarmann > > Jack > > > On 9/20/07, Jack Gilbert <gilbertjack(a)gmail.com> wrote: > Dan, > > I have another one with the same problem, however, the compressed > file is still 14Mg. > > Let me know what to do. > > Jack > > > On 9/20/07, Jack Gilbert <gilbertjack(a)gmail.com > wrote: > Dan, > > Have also been having problems uploading this file (I have attached > it compressed). It says each time that it cannot detect FASTA in > the file. Obviously they are FASTA..so whats going on? > > Jack > > > On 9/20/07, Jack Gilbert <gilbertjack(a)gmail.com > wrote: > Dan, > > Virtually all of them would be good to delete except two. > Therefore, please could you remove: > > 183, 184, 186, 187, 188, 189, 190, 191, 192, 195, 198, 199, 200, 201 > > These were all preliminary runs and they are getting things > confused now. will upload the new ones now. > > Cheers > > Jack > > On 9/19/07, Daniel Paarmann <paarmann(a)mcs.anl.gov > wrote: > Jack, > > > the submission of sequences to SEED appears to of changed. It is > > coming up with a genome centric submission which seems to have no > > place for metagenomic samples. > > thank you for notifying us about this. It has been corrected and the > metagenome RAST server shows the correct upload page again. > > Unfortunately, due to this configuration error on our side, your jobs > 202-205 have been created as regular RAST jobs. I had to delete them > and would ask you to upload them again. My apologies for any > inconvenience this may cause. > 202 All All All > 203 Bag1 13thMay cDNA > 204 Bag6 19thMay cDNA > 205 Bag1 19thMay cDNA > > > Also, quick question regarding the taxanomic analysis, the > > taxanomic analysis data keeps on changing - at least everytime I > > look at it I see problems or things that have changed. Is this in > > continuous flux? > > The taxonomic analysis and the databases used for it have not been > changed since the server release. We appreciate any feedback about > inconsistencies and will investigate it immediately. Could you be > more specific for what metagenome (job) you noticed differences? What > kind of changes did you observe? > > Regarding your question how to delete jobs, this action is currently > not supported by the RAST web service, but I will suggest it for our > next release. For now, if you send my the job numbers and metagenome > ids, I'll gladly delete them for you. > > > Best regards, > Daniel Paarmann > > > > -- > Dr. Jack A. Gilbert B.Sc. Ph.D. > Molecular Ecologist > Plymouth Marine Laboratory, > Prospect Place, The Hoe, > Plymouth, PL1 3DH, UK. > Work Tel: +44 (0)1752 633 416 > Home Tel: +44 (0)1752 255 783 > Mobile Tel: +44 (0)7764 190 147 > Email: gilbertjack(a)gmail.com > > > > -- > Dr. Jack A. Gilbert B.Sc. Ph.D. > Molecular Ecologist > Plymouth Marine Laboratory, > Prospect Place, The Hoe, > Plymouth, PL1 3DH, UK. > Work Tel: +44 (0)1752 633 416 > Home Tel: +44 (0)1752 255 783 > Mobile Tel: +44 (0)7764 190 147 > Email: gilbertjack(a)gmail.com > > > > > -- > Dr. Jack A. Gilbert B.Sc. Ph.D. > Molecular Ecologist > Plymouth Marine Laboratory, > Prospect Place, The Hoe, > Plymouth, PL1 3DH, UK. > Work Tel: +44 (0)1752 633 416 > Home Tel: +44 (0)1752 255 783 > Mobile Tel: +44 (0)7764 190 147 > Email: gilbertjack(a)gmail.com > > > > -- > Dr. Jack A. Gilbert B.Sc. Ph.D. > Molecular Ecologist > Plymouth Marine Laboratory, > Prospect Place, The Hoe, > Plymouth, PL1 3DH, UK. > Work Tel: +44 (0)1752 633 416 > Home Tel: +44 (0)1752 255 783 > Mobile Tel: +44 (0)7764 190 147 > Email: gilbertjack(a)gmail.com

1 0

mg-rast Re: Questions about mg rast
by Daniel Paarmann 20 Sep '07

20 Sep '07

Jack, the upload page fails to detect the FASTA file inside your zip archive because it currently uses a very simple mechanism based on the file extension. I've filed a bug for this on our todo list to make it more robust. For now you can rename your FASTA file to lower case *.fna, *.fa, or *.fasta at the end, optionally re-compress it and upload it again. If you run into any other problems with the uploads, please let me know. Best regards, Daniel On Sep 20, 2007, at 6:17 AM, Jack Gilbert wrote: > Dan, > > Have also been having problems uploading this file (I have attached > it compressed). It says each time that it cannot detect FASTA in > the file. Obviously they are FASTA..so whats going on? > > Jack >

1 0

mg-rast Re: [rast] [Fwd: ] rast server login
by enavarro 20 Sep '07

20 Sep '07

Hello I received the login for elisabeth.navarro2(a)wanadoo.fr (Sept. 20th) this Afternoon Thanks Elisabeth Navarro Le 20 sept. 07 ï¿½ 16:11, Folker Meyer a ï¿½crit : > Hello Elisabeth, > > we have several requests in your name, some with an ec-lyon email > address and 2 with another address. All of the above have been > ok'ed by us. > > There should now be 4 login in your name: > > - Isabelle Navarro, inavarro, navarro(a)univ-lyon1.fr (Sept. 7th) > - Elisabeth Navarro, enavarro, navarro(a)ec-lyon.fr (Sept. 11th) > - Elisabeth Navarro, enavarro1, Elisabeth.Navarro(a)ec-lyon.fr (Sept. > 11th) > - Isabelle Navarro, inavarro1, elisabeth.navarro2(a)wanadoo.fr (Sept. > 20th) > > Our server did send an email for each login listed above. > > Maybe your SPAM filter is filtering out the emails our server is > sending? Please check your SPAM folder. > > > > Best regards, > Folker Meyer > > Begin forwarded message: > >> From: William Mihalo <wmihalo(a)ci.uchicago.edu> >> Date: September 20, 2007 8:48:13 AM CDT >> To: rast(a)mcs.anl.gov >> Subject: [rast] [Fwd: ] rast server login >> >> forwarded from info(a)theseed.org >> >> From: Isabelle Navarro <Elisabeth.Navarro(a)ec-lyon.fr> >> Date: September 20, 2007 3:42:31 AM CDT >> To: info(a)theseed.org >> >> >> Hello, >> >> Last week, I tried to create an account in order to used the >> Metagenomics RAST Server. Nevertheless, I never received the login >> and the password. I would like to uses SEED in order to analyze >> metagenomic data and I really appreciate if you coud give a login >> and a password >> >> >> Thanks, sincerely yours >> >> Elisabeth Navarro >> >> >> Elisabeth.Navarro(a)ec-lyon.fr >> Equipe ï¿½ Gï¿½nomique Microbienne Environnementale ï¿½ (Environmental >> Microbial Genomics Group ) >> UMR CNRS 5005, Laboratoire Ampï¿½re, Ecole Centrale de Lyon , 36 >> avenue Guy de Collongue, 69134 Ecully Cedex, France >> >> www.GenomEnviron.org >

1 0

mg-rast Fwd: [rast] [Fwd: ] rast server login
by Folker Meyer 20 Sep '07

20 Sep '07

Hello Elisabeth, we have several requests in your name, some with an ec-lyon email address and 2 with another address. All of the above have been ok'ed by us. There should now be 4 login in your name: - Isabelle Navarro, inavarro, navarro(a)univ-lyon1.fr (Sept. 7th) - Elisabeth Navarro, enavarro, navarro(a)ec-lyon.fr (Sept. 11th) - Elisabeth Navarro, enavarro1, Elisabeth.Navarro(a)ec-lyon.fr (Sept. 11th) - Isabelle Navarro, inavarro1, elisabeth.navarro2(a)wanadoo.fr (Sept. 20th) Our server did send an email for each login listed above. Maybe your SPAM filter is filtering out the emails our server is sending? Please check your SPAM folder. Best regards, Folker Meyer Begin forwarded message: > From: William Mihalo <wmihalo(a)ci.uchicago.edu> > Date: September 20, 2007 8:48:13 AM CDT > To: rast(a)mcs.anl.gov > Subject: [rast] [Fwd: ] rast server login > > forwarded from info(a)theseed.org > > From: Isabelle Navarro <Elisabeth.Navarro(a)ec-lyon.fr> > Date: September 20, 2007 3:42:31 AM CDT > To: info(a)theseed.org > > > Hello, > > Last week, I tried to create an account in order to used the > Metagenomics RAST Server. Nevertheless, I never received the login > and the password. I would like to uses SEED in order to analyze > metagenomic data and I really appreciate if you coud give a login > and a password > > > Thanks, sincerely yours > > Elisabeth Navarro > > > Elisabeth.Navarro(a)ec-lyon.fr > Equipe ï¿½ Gï¿½nomique Microbienne Environnementale ï¿½ (Environmental > Microbial Genomics Group ) > UMR CNRS 5005, Laboratoire Ampï¿½re, Ecole Centrale de Lyon , 36 > avenue Guy de Collongue, 69134 Ecully Cedex, France > > www.GenomEnviron.org

1 0