Selected real-source examples
These artifacts are generated from unchanged retained sources by explicit selection. They are distinct from deliberately malformed and biologically altered fixtures.
actinorhodin.gff3 is reproduced by just bgc-report; see the
BGC exercise for its scientific question,
selection rule, full source and limits on round-trip claims.
populus-v4.1-chr01-200-genes.gene_exons.gff3 holds the first 200 genes on Chr01 of Populus
trichocarpa v4.1 (Phytozome-533), with all their mRNA, exon, CDS and UTR rows unchanged, the
source's three directives, and four provenance lines: the three every derived file carries and a
# cite: line with the citations below, so the file names them itself. scripts/populus_excerpt.py writes it
from Ptrichocarpa_533_v4.1.gene_exons.gff3.gz (JGI file 5d94dc9fc0d65a87debccfce, md5
3fdcfdd5af213c5f4dde0c01f2cdbfda), which needs a JGI login and is too large to vendor, and checks
that md5 first. Phytozome lists this genome as public with citation required. Cite Tuskan et al.
2006, Science 313:1596 (doi:10.1126/science.1128691), and Goodstein et al. 2012, Nucleic Acids Res
40:D1178. JGI's legacy data policy asks for this acknowledgment (whether that policy governs this
genome is undetermined, see docs/restricted-sources.md): "These sequence data were
produced by the US Department of Energy Joint Genome Institute http://www.jgi.doe.gov/ in
collaboration with the user community."
populus-v4.1-chr01-200-genes.annotation_info.txt holds the 293 rows of
Ptrichocarpa_533_v4.1.annotation_info.txt (JGI file 5d94dc9ec0d65a87debccfcc, md5
aaf04aa3515d96748d1229d36ac7dfe4) for the 200 genes of the gene_exons excerpt, unchanged and in
table order, after the header and the same four provenance lines. scripts/populus_excerpt.py --table
writes it, checking that md5 first. Two of its rows have a peptideName naming another locus's
transcript. The header's KOG and ec labels are swapped, as in the source. The same citations apply.