Validation of metabolights configurations.

[pconesa]

Hi Eammon.

Our isatab files based on our configurations do not pass the validation.

Our configuration has "MS Assay Name" as a field, but the validator, somehow is expecting "Assay Name".

I have created a isatab file with ISACreator v1.5.0 and our configurations. I have used validator 1.4 and the same configurations (ours).

Log file says: Message: Invalid ISA-TAB: unexpected field MS Assay Name found in section generic_assay_pipeline

Section.getField("MS Assay Name") returns null.

So, RowSectionParser.parseHeaders throws an exception: throw new TabValidationException(i18n.msg("unexpected_field_in_section_error", header, sectionId));

(SchemaNode) Schema.children has "Assay Name" instead of "MS Assay Name".

Any idea about what is going on here?

[eamonnmag]

@proccaserra is now looking at this.

[pconesa]

I have been able to validate it moving the "Data Transformation Name" field just before the last one ("Metabolite Assigment File"). By default (when I create a new assay) it is placed after "Normalization Name".

Does this help? (I guess you are already aware of it) Is this, again, a wrong configuration problem?

Cheers!

[proccaserra]

HI Pablo

I was actually writing this to you now! sorry it was our Xmas lunch today and had way too much wine hence the delay in getting back to you. Regarding your second question: 'is it a bad configuration?' well, I am running so more tests to make a complete call. the thing is that ISAconfigurator, ISAvalidator and ISA specification were developped in parallel when the specifications were not yet entirely solidified. Therefore, Marco and Eamonn had to do some blind flying. To compound that fact, It is seems I have made wrong assumptions about how both ISAcreator and ISAvalidator would interpret the configurations so we are identifying/solving those inconsistencies as we go along. I expect to cure the problem by moving a Protocol REF from its current position to place it between Normalization Name and Data Transformation Name. I will run a series of tests in order to confirm this is good enough, I will also need to validate all previous experiments in order to ensure backward compatibility

Again, apologies for the problems this is causing

Best wishes

Philippe

On 14/12/2011 17:44, Pablo Conesa wrote:

I have been able to validate it moving the "Data Transformation Name" field just before the last one ("Metabolite Assigment File"). By default (when I create a new assay) it is placed after "Normalization Name".

Does this help? (I guess you are already aware of it) Is this, again, a wrong configuration problem?

Cheers!

---

Reply to this email directly or view it on GitHub: https://github.com/ISA-tools/ISAvalidator-ISAconverter-BIImanager/issues/35#issuecomment-3145150

[pconesa]

Ok, it is good to know you have the problem isolated and "surrounded".

No worries, Philippe, Xmas lunch is more important, and more if you had fun and wine....

Many thanks for your response....so far, we can know continue working as now it's possible to upload to metabolights.

All the best, Pablo

[eamonnmag]

Is this problem now solved?

[pconesa]

Theorically yes! I guess we can reopen it again in case, there's something missing. I hope to try it tomorrow.

EDITED: Sorry, Eamonn. I thought this was referring to the model-index problem. I am afraid it is not solved. Unless, we are creating wrong, non-sense ISATab files (which is likely to happen).

But today I have tried to validate Ken's file (as dummy as mine's) and it didn't pass the validation (using official validator 1.4 for MAC). ISacreator validation process validates!.

On 10/01/2012 17:08, Eamonn Maguire wrote:

Is this problem now solved?

---

Reply to this email directly or view it on GitHub: https://github.com/ISA-tools/ISAvalidator-ISAconverter-BIImanager/issues/35#issuecomment-3432959

Pablo Conesa Mingo Skype: pconesa Phone: 01223 494596 EMBL - European Bioinformatics Institute. Wellcome Trust Genome Campus, Hinxton, Cambridge, CB10 1SD, UK

[eamonnmag]

Closed :)

[pconesa]

Hi Eamonn!! Almost getting there!!!

Everything works fine apart from the "DataLocationManager.getDataLocationLink". I am getting an empty string. Any ideas why? Datalocations file looks correct and files are copied to their proper locations after an upload...

[eamonnmag]

What are the parameters you are supplying to the method?

[pconesa]

I haven't changed them:

assay.getMeasurement().getName(), assay.getTechnologyName(), assay.getStudy().getObfuscationCode(), AnnotationTypes.GENERIC_DATA_FILE_PATH

[eamonnmag]

For the sake of the conversation thread, this is now working :)