OHDSI / MedlineXmlToDatabase

A command line Java application for parsing MEDLINE XML files and inserting the data into a relational database
Apache License 2.0
19 stars 11 forks source link

where is cited pmids? #8

Closed lemonysoft closed 6 years ago

lemonysoft commented 6 years ago

Hi, I have downloaded .tar.gz files and converted all by MedlineXMLToDatabase tool. But I can't find PMIDs in commentscorrectslist table although they exist in original XML files.

Do you know where I can find them in which table? Thanks.

schuemie commented 6 years ago

I have 152,671,021 records in the table medcit_commentscorrectionslist_commentscorrections. Are there specific PMIDs you were expecting to see there that are missing?

lemonysoft commented 6 years ago

No. medcit_commentscorrectionslist_commentscorrections table has these fields: order, note, pmid, pmid_version, refsource and reftype. But cited reference PMIDs column is missing in the table. However original XML file has it as shown in following example.

<PubmedArticle>
    <MedlineCitation Status="MEDLINE" Owner="NLM">
        <PMID Version="1">23423664</PMID>
.
.
<CommentsCorrectionsList>
            <CommentsCorrections RefType="Cites">
                <RefSource>J Pers Assess. 1985 Feb;49(1):71-5</RefSource>
                <PMID Version="1">16367493</PMID>
            </CommentsCorrections>
            <CommentsCorrections RefType="Cites">
                <RefSource>J Pain Symptom Manage. 2008 Feb;35(2):203-13</RefSource>
                <PMID Version="1">18158230</PMID>
            </CommentsCorrections>
schuemie commented 6 years ago

Ach crivens! The PMID field in that table gets overwritten with the PMID corresponding to the parent citation.

The code is easily fixed: I am now simply renaming such fields to 'Other_PMID". However, it will require both the analysis and parsing to be completely redone.

lemonysoft commented 6 years ago

Are you going to publish a new release? Awaiting for you :)

schuemie commented 6 years ago

Yep, here's the new release.

lemonysoft commented 6 years ago

Thanks a lot, really appreciated. You did a great job for us.