Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[bioontology-support] *Second Request* - The annotations in the output report for the Annotator do Not always match the Annotations in the JSON file...

Al Shapiro als371 at yahoo.com
Thu Oct 24 17:34:28 PDT 2013


Ray,

When you have some time, can you send me anexample of the batch call passing all classes, so that I can then get access to all the preferred names, because it is the "preferred names" that I consider to be the multiple annotations that I am looking for, for my project....

In other words, how does one do this process!

Thank you for your quick response because I know you guys are swamped with support e-mails...

Al Shapiro 







On Thursday, October 24, 2013 4:40 PM, Ray Fergerson <ray.fergerson at stanford.edu> wrote:
 
Sorry about the delay. I think that I understand the question. I think that the service and the file are working correctly. I believe that the problem is that it is difficult to see the connection between what it displayed in the table and the contents of the json file. Perhaps this will help:
 
The first column in the table is the preferred name of the matched class. This name does not necessarily appear anywhere in the json output. We make another call to get the preferred name for all matched classes (a batch call passing all classes and getting back all preferred names). The “annotations text” in the result may or may not be the same as the preferred name. For example if the match was on a synonym then it will not be the same. Likewise if the match was a hierarchy or mapping match then it will not be the same.
 
Ray
 
From:Al Shapiro [mailto:als371 at yahoo.com] 
Sent: Thursday, October 24, 2013 12:20 PM
To: Ray Fergerson; bioontology-support at lists.stanford.edu; support at bioontology.org
Cc: Al; Mona Salem
Subject: *Second Request* - The annotations in the output report for the Annotator do Not always match the Annotations in the JSON file...
 
Ray and support personnel,
 
I know you guys are busy but I would appreciate some feedback on the following BioPortal Annotator problem that I have outlined below the following:
 
a) Acknowledgement that it is or it is Not a problem;
 
b) If it is Not a problem, how can I locate (via the JSON output file) the matching annotations that display in the Annotation report;
 
c) If it is deemed a problem, approximately how long will it take for you guys to fix it.
 
Thanks again - see details about this problem below,
 
Al Shapiro
 
On Monday, October 21, 2013 5:17 PM, Al Shapiro <als371 at yahoo.com> wrote:
Hi there again,
 
The annotations in the output report for the Annotator do Not always match the Annotations in the JSON file...
 
The annotations from the output report in column 1 do Not always match the "annotations text" in the JSON file...
 
Note: Over 60% of the Annotations on the displayed report do Not match what is in the JSON file...
 
entry 1) column 1 of the report = NCI CTEP SDC Melanoma Sub-Category Terminology;
and the first "annotations text" entry in the JSON file  =  "MELANOMA"
 
entry 2) column 1 of the report = Melanoma Pathway; 
and the second "annotations text" entry in the JSON file  =  "MELANOMA"
 
entry 3) column  1 of the report = Melanoma; 
and the third "annotations text" entry in the JSON file  =  "MELANOMA"
 
entry 4) column  1 of the report = Mouse Melanoma; 
and the fourth "annotations text" entry in the JSON file  =  "MELANOMA"
 
entry 5) column  1 of the report = Malignant Neoplasm; 
and the fifth "annotations text" entry in the JSON file  =  "MALIGNANT TUMOR"
 
entry 6) column  1 of the report = Malignant Paraganglioma; 
and the sixth "annotations text" entry in the JSON file  =  "MALIGNANT"
 
entry 7) column  1 of the report = Anaplastic Oligodendroglioma; 
and the seventh "annotations text" entry in the JSON file  =  "MALIGNANT"
 
entry 8) column  1 of the report = MALIGNANT; 
and the eighth "annotations text" entry in the JSON file  =  "MALIGNANT"
 
entry 9) column  1 of the report = Neoplasm; 
and the ninth "annotations text" entry in the JSON file  =  "Tumor"
 
entry 10) column  1 of the report = Tumor Mass; 
and the tenth "annotations text" entry in the JSON file  =  "Tumor"
 
entry 11) column  1 of the report = Melanocyte; 
and the eleventh "annotations text" entry in the JSON file  =  "MELANOCYTES"
 
entry 12) column  1 of the report = Murine Melanocytes; 
and the twelfth "annotations text" entry in the JSON file  =  "MELANOCYTES"
 
entry 13) column  1 of the report = Present; 
and the thirteenth "annotations text" entry in the JSON file  =  "FOUND"
 
entry 14) column  1 of the report = Skin; 
and the fourteenth "annotations text" entry in the JSON file  =  "SKIN"
 
entry 15) column  1 of the report = Murine Skin; 
and the fifteenth "annotations text" entry in the JSON file  =  "SKIN"
 
entry 16) column  1 of the report = Mouse Skin; 
and the sixteenth "annotations text" entry in the JSON file  =  "SKIN"
 
entry 17) column  1 of the report = Murine Intestines; 
and the seventeenth "annotations text" entry in the JSON file  =  "BOWEL"
 
entry 18) column  1 of the report = Intestine; 
and the eighteenth "annotations text" entry in the JSON file  =  "BOWEL"
 
entry 19) column  1 of the report = Murine Eye; 
and the nineteenth "annotations text" entry in the JSON file  =  "EYE"
 
entry 20) column  1 of the report = Eye; 
and the twentieth "annotations text" entry in the JSON file  =  "EYE"
 
entry 21) column  1 of the report = Mouse Eye; 
and the twenty first "annotations text" entry in the JSON file  =  "EYE"
 
Do the following to reproduce the above results:
 
1) Go to the  BioPortal Manual Annotator page;
2) Click on "insert sample text"; [Melanoma is a malignant tumor of melanocytes which are found predominantly in skin but also in the bowel and the eye.]
3) In the "Select Ontologies" box enter the National Cancer Institute Thesaurus (NCIT) ontology;
4) Click on "Get Annotations" button;
5) After the Annotations report displays on the screen, Click on Format Results As:     JSON link;
6) Examine results from both outputs.
 
Note:My company is developing an application system that needs to Annotate the Text contained in the biological software product abstracts located at The G6G Directory of Omics and Intelligent Software web-site: http://g6g-softwaredirectory.com/
 
Any insight would be greatly appreciated,
 
Thank you,
 
Al Shapiro
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.stanford.edu/pipermail/bioontology-support/attachments/20131024/767a89ac/attachment-0003.html>


More information about the bioontology-support mailing list