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] don't use subtree search with maxnumhits

Ray Fergerson ray.fergerson at
Fri Aug 30 17:22:44 PDT 2013



You are correct that this combination is handled badly. First the returned
results are truncated and then the subtree pruning happens. If we change
the order though the call we become very slow. Subtree pruning is a slow
operation in the current system. We hope that it will eventually become
quicker with our new backend that we are working on.


Suggest, for the moment, a hack to just increase the count of returned
results and let the extras just be pruned away. You might experiment with
100 or 1000. This will be slower but if you are doing in in batch it may
still be ok. This is obviously not a great fix but it may be better than




From: bioontology-support-bounces at
[mailto:bioontology-support-bounces at] On Behalf Of Lee M
Sent: Wednesday, August 28, 2013 7:36 AM
To: support
Subject: [bioontology-support] don't use subtree search with maxnumhits


I was trying to use the REST API to search chemo regimen in NCI Thesaurus
by setting the subtreeconceptid to the Chemotherapy Regimen term (C62634).

When I searched for ATRA, I was shocked that there were no hits:
0 Hits!

Upon investigation, it turns out that setting maxnumhits was the issue.
Maybe it is doing the subtree filter AFTER taking the top 15 hits?!
90 Hits!

Could you please confirm my understanding?  Is it the intended behavior?


Lee Surprenant
IBM Emerging Technologies | jStart Team
lmsurpre at | (919) 543-8919

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the bioontology-support mailing list