The recently added LactoPanGEM (816 NCBI) and Ecoli_panGEMs (330 NCBI + 1972 PATRIC) lack genome assembly data. Generally, both model files (ex. sbml, json) and the corresponding genome assembly files (ex. genomic.gbff[.gz]) are needed, and the latter provide gene locus tag, sequence, location info on the gene page. To support quick ingestion, two constraints in the loader were relaxed: a new patric_assembly: type was added, and ncbi_assembly: now also accepts entries without a local genome assembly file.
Even though the model, genome, gene pages all can be correctly rendered lacking assembly data, two problems remain to enrich gene page info:
-
The assembly files which the models were based on are missing. They should be obtained from the author, since these curated panGEMs are not derived from public NCBI annotations.
-
For curated panGEMs, assembly files usually use annotate tool specific locus tag naming convention, which could differ from the NCBI RefSeq locus tags largely used in the BiGGr database. Even with the original assembly files in hand, this mismatch still degrades the value of the gene page. Mapping the locus tags to the NCBI RefSeq locus is also a problem to solve.
The recently added LactoPanGEM (816 NCBI) and Ecoli_panGEMs (330 NCBI + 1972 PATRIC) lack genome assembly data. Generally, both model files (ex. sbml, json) and the corresponding genome assembly files (ex. genomic.gbff[.gz]) are needed, and the latter provide gene locus tag, sequence, location info on the gene page. To support quick ingestion, two constraints in the loader were relaxed: a new patric_assembly: type was added, and ncbi_assembly: now also accepts entries without a local genome assembly file.
Even though the model, genome, gene pages all can be correctly rendered lacking assembly data, two problems remain to enrich gene page info:
The assembly files which the models were based on are missing. They should be obtained from the author, since these curated panGEMs are not derived from public NCBI annotations.
For curated panGEMs, assembly files usually use annotate tool specific locus tag naming convention, which could differ from the NCBI RefSeq locus tags largely used in the BiGGr database. Even with the original assembly files in hand, this mismatch still degrades the value of the gene page. Mapping the locus tags to the NCBI RefSeq locus is also a problem to solve.