|Version 1 (modified by markjschreiber, 6 years ago)|
Working with BioJava?-live SVN version 4723.
Sequence case is not preserved. Line length varies (default is 80 cpl). ===Read-Write-Read=== Succeeds!
Feature qualifier order is not preserved. Because NCBI Taxonomy is referenced from memory or database if the version used doesn't match the version that was used to construct the record then minor differences appear. For example the common name of Arabidopsis changed from thale cress to mouse ear cress. ===Read-Write-Read=== Success!
Not supported (INSD is).
* BioJava? inlcudes the Reference_position tag. NCBI doesn't unless it is not 1..1
* There are other examples of this redundancy. I think if this doesn't break the dtd then it doesn't matter. * Qualifiers order is not preserved. I don't think this matters. ===Read-Write-Read=== Success!
No major issues
- Version in date is not correct on output.
- Two XX lines after references.
- Feature qualifiers out of order.
- BioSQL cannot store more than one database reference for a single publication,
eg Pubmed and medline Id and DOI.
- We are putting 'and' before the last author.
- We loose the copyright statement.
===Read-Write-Read=== Cannot read back in:
<code> Format_object=org.biojavax.bio.seq.io.UniProtFormat? Accession=Q43385 Id= Comments= Parse_block=OS Arabidopsis thaliana (Mouse-ear cress) (Arabidopsis thaliana (L.) Heynh.).OC Eukaryota; Viridiplantae; Streptophyta; Embryophyta; Tracheophyta;OS Spermatophyta; Magnoliophyta; eudicotyledons; core eudicotyledons; rosids; eurosids II; Brassicales; Brassicaceae; Arabidopsis.OX NCBI_TaxID=3702; Stack trace follows ....
Caused by: java.lang.IllegalArgumentException?: NCBI taxonomy names cannot embed new lines - at:74, in name: <Spermatophyta; Magnoliophyta; eudicotyledons; core eudicotyledons; rosids; eurosids II; Brassicales; Brassicaceae; Arabidopsis>
at org.biojavax.bio.taxa.SimpleNCBITaxonName.<init>(SimpleNCBITaxonName.java:47) at org.biojavax.bio.taxa.SimpleNCBITaxon.addName(SimpleNCBITaxon.java:148) at org.biojavax.bio.seq.io.UniProtFormat?.readRichSequence(UniProtFormat?.java:339) ... 4 more
- Missing the namespace and the version from entry tag
- Don't write editor list if there are no editors
- Reference tag not correctly constructed.
- When writing Uniprot XML it would be illegal to use anything other than
Swiss-Prot or TrEMBL as the Namespace