You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@ctakes.apache.org by "MAZE, Christian" <ch...@sogeti.com> on 2017/03/23 13:35:34 UTC
detection of chemical/drug entities
Hi,
I am using the v3.2.2 ctakes.
I copied the v3.2.1.1 resources.
I used another tool for extracting chemical/drug entities and when processing my input text, I got the following results:
1092 1107 nucleotide 1517
1165 1178 aspartic acid
1184 1191 glycine
1195 1204 codon 506
1254 1257 Asp
My input test text is :
A novel missense mutation Asp506Gly in Exon 13 of the F11 gene in an asymptomatic Korean woman with mild factor XI deficiency. Factor XI (FXI) deficiency is a rare autosomal recessive coagulation disorder most commonly found in Ashkenazi and Iraqi Jews, but it is also found in other ethnic groups. It is a trauma or surgery-related bleeding disorder, but spontaneous bleeding is rarely seen. The clinical manifestation of bleeding in FXI deficiency cases is variable and seems to poorly correlate with plasma FXI levels. The molecular pathology of FXI deficiency is mutation in the F11 gene on the chromosome band 4q35. We report a novel mutation of the F11 gene in an 18-year-old asymptomatic Korean woman with mild FXI deficiency. Pre-operative laboratory screen tests for lipoma on her back revealed slightly prolonged activated partial thromboplastin time (45.2 sec; reference range, 23.2-39.4 sec). Her FXI activity (35%) was slightly lower than the normal FXI activity (reference range, 50-150%). Direct sequence analysis of the F11 gene revealed a heterozygous A to G substitution in nucleotide 1517 (c.1517A>G) of exon 13, resulting in the substitution of aspartic acid with glycine in codon 506 (p.Asp506Gly). To the best of our knowledge, the Asp506Gly is a novel missense mutation, and this is the first genetically confirmed case of mild FXI deficiency in Korea.
When I start the same treatment using ctakes and the AggregatePlaintextFastUMLSProcessor annotator file.
No drug entity is detected inside drugner leaf (
[cid:image001.png@01D2A3CE.242F5D40]
For example I would have thought that nucleotide be classified in the drugner leaf.
Please could you correct me if I missed anything ?
Christian
This message contains information that may be privileged or confidential and is the property of the Capgemini Group. It is intended only for the person to whom it is addressed. If you are not the intended recipient, you are not authorized to read, print, retain, copy, disseminate, distribute, or use this message or any part thereof. If you receive this message in error, please notify the sender immediately and delete all copies of this message.
Re: detection of chemical/drug entities
Posted by James Masanz <ma...@gmail.com>.
Sorry, parts of my previous reply was incorrect.
I forgot that as of 3.1, the annotations made by the drugner component are
converted into org.apache.ctakes.typesystem.type.textsem.MedicationMention
annotations. So the AggregatePlaintextFastUMLSProcessor is working as
expected -- instead of looking for medications as annotations of type
org.apache.ctakes.drugner.type.* use org.apache.ctakes.typesystem.
type.textsem.MedicationMention.
As I mentioned before, whether nucleotide is classified as a medication
depends on the dictionary used. It appears nucleotide is not within the
default dictionary as a medication. However I do see annotations being
created for glycine and aspartic acid.
org.apache.ctakes.typesystem.type.textsem.MedicationMention is found under
org.apache.ctakes.typesystem.type.textsem.EventMention which is found under
org.apache.ctakes.typesystem.type.textsem.IdentifiedAnnotation
To get attributes such as strength, dose, etc, a MedicationMention
annotation has attributes such as medicationStrength which can reference an
annotation of type MedicationStrengthModifier
Regards,
-- James
On Thu, Mar 23, 2017 at 9:35 AM, MAZE, Christian <ch...@sogeti.com>
wrote:
> Hi,
>
>
>
> I am using the v3.2.2 ctakes.
>
> I copied the v3.2.1.1 resources.
>
>
>
> I used another tool for extracting chemical/drug entities and when
> processing my input text, I got the following results:
>
> 1092 1107 nucleotide 1517
>
> 1165 1178 aspartic acid
>
> 1184 1191 glycine
>
> 1195 1204 codon 506
>
> 1254 1257 Asp
>
>
>
> My input test text is :
>
> A novel missense mutation Asp506Gly in Exon 13 of the F11 gene in an
> asymptomatic Korean woman with mild factor XI deficiency. Factor XI (FXI)
> deficiency is a rare autosomal recessive coagulation disorder most commonly
> found in Ashkenazi and Iraqi Jews, but it is also found in other ethnic
> groups. It is a trauma or surgery-related bleeding disorder, but
> spontaneous bleeding is rarely seen. The clinical manifestation of bleeding
> in FXI deficiency cases is variable and seems to poorly correlate with
> plasma FXI levels. The molecular pathology of FXI deficiency is mutation in
> the F11 gene on the chromosome band 4q35. We report a novel mutation of the
> F11 gene in an 18-year-old asymptomatic Korean woman with mild FXI
> deficiency. Pre-operative laboratory screen tests for lipoma on her back
> revealed slightly prolonged activated partial thromboplastin time (45.2
> sec; reference range, 23.2-39.4 sec). Her FXI activity (35%) was slightly
> lower than the normal FXI activity (reference range, 50-150%). Direct
> sequence analysis of the F11 gene revealed a heterozygous A to G
> substitution in nucleotide 1517 (c.1517A>G) of exon 13, resulting in the
> substitution of aspartic acid with glycine in codon 506 (p.Asp506Gly). To
> the best of our knowledge, the Asp506Gly is a novel missense mutation, and
> this is the first genetically confirmed case of mild FXI deficiency in
> Korea.
>
>
>
> When I start the same treatment using ctakes and the
> AggregatePlaintextFastUMLSProcessor annotator file.
>
>
>
> No drug entity is detected inside drugner leaf (
>
> For example I would have thought that nucleotide be classified in the
> drugner leaf.
>
>
>
> Please could you correct me if I missed anything ?
>
>
>
> Christian
>
>
> This message contains information that may be privileged or confidential
> and is the property of the Capgemini Group. It is intended only for the
> person to whom it is addressed. If you are not the intended recipient, you
> are not authorized to read, print, retain, copy, disseminate, distribute,
> or use this message or any part thereof. If you receive this message in
> error, please notify the sender immediately and delete all copies of this
> message.
>
Re: detection of chemical/drug entities
Posted by James Masanz <ma...@gmail.com>.
Whether nucleotide is classified as a medication depends on the dictionary
used. It appears nucleotide is not within the default dictionary as a
medication
In order to have drugner annotations,
org.apache.ctakes.drugner.ae.DrugMentionAnnotator
needs to be in the pipeline.
(When using xml descriptors, that's
desc\ctakes-drug-ner\desc\analysis_engine\DrugMentionAnnotator.xml)
However, I see that annotator listed in the AggregatePlaintextFastUMLSProcessor
pipeline. I'll take a look at why the drugner annotations are not being
created.
When I use AggregatePlaintextFastUMLSProcessor in 3.2.2, medications *are*
getting annotated - but they are being annotated as
org.apache.ctakes.typesystem.type.textsem.MedicationEventMention
which is found under
org.apache.ctakes.typesystem.type.textsem.EventMention
hope this helps
-- James
On Thu, Mar 23, 2017 at 9:35 AM, MAZE, Christian <ch...@sogeti.com>
wrote:
> Hi,
>
>
>
> I am using the v3.2.2 ctakes.
>
> I copied the v3.2.1.1 resources.
>
>
>
> I used another tool for extracting chemical/drug entities and when
> processing my input text, I got the following results:
>
> 1092 1107 nucleotide 1517
>
> 1165 1178 aspartic acid
>
> 1184 1191 glycine
>
> 1195 1204 codon 506
>
> 1254 1257 Asp
>
>
>
> My input test text is :
>
> A novel missense mutation Asp506Gly in Exon 13 of the F11 gene in an
> asymptomatic Korean woman with mild factor XI deficiency. Factor XI (FXI)
> deficiency is a rare autosomal recessive coagulation disorder most commonly
> found in Ashkenazi and Iraqi Jews, but it is also found in other ethnic
> groups. It is a trauma or surgery-related bleeding disorder, but
> spontaneous bleeding is rarely seen. The clinical manifestation of bleeding
> in FXI deficiency cases is variable and seems to poorly correlate with
> plasma FXI levels. The molecular pathology of FXI deficiency is mutation in
> the F11 gene on the chromosome band 4q35. We report a novel mutation of the
> F11 gene in an 18-year-old asymptomatic Korean woman with mild FXI
> deficiency. Pre-operative laboratory screen tests for lipoma on her back
> revealed slightly prolonged activated partial thromboplastin time (45.2
> sec; reference range, 23.2-39.4 sec). Her FXI activity (35%) was slightly
> lower than the normal FXI activity (reference range, 50-150%). Direct
> sequence analysis of the F11 gene revealed a heterozygous A to G
> substitution in nucleotide 1517 (c.1517A>G) of exon 13, resulting in the
> substitution of aspartic acid with glycine in codon 506 (p.Asp506Gly). To
> the best of our knowledge, the Asp506Gly is a novel missense mutation, and
> this is the first genetically confirmed case of mild FXI deficiency in
> Korea.
>
>
>
> When I start the same treatment using ctakes and the
> AggregatePlaintextFastUMLSProcessor annotator file.
>
>
>
> No drug entity is detected inside drugner leaf (
>
> For example I would have thought that nucleotide be classified in the
> drugner leaf.
>
>
>
> Please could you correct me if I missed anything ?
>
>
>
> Christian
>
>
> This message contains information that may be privileged or confidential
> and is the property of the Capgemini Group. It is intended only for the
> person to whom it is addressed. If you are not the intended recipient, you
> are not authorized to read, print, retain, copy, disseminate, distribute,
> or use this message or any part thereof. If you receive this message in
> error, please notify the sender immediately and delete all copies of this
> message.
>