Treat biological databases as infrastructure, not projects www.nature.com/articles/d41...

Somewhat behind schedule (technical reasons) we have uploaded principal isoforms for @gencodegenes.bsky.social G49 (human) and M38 (mouse) versions, corresponding to @ensembl.org 115. 239 human genes have at least one manual APPRIS annotation and 133 are annotated with Principal:M transcripts.

APPRIS currently has an expired certificate. Unfortunately, the IT department is particularly short-staffed (for reasons beyond our control) and we cannot estimate when the problem will be fixed.

More than 100 dual coding regions have evidence for selection constraints in both reading frames https://www.biorxiv.org/content/10.64898/2026.05.28.727796v1

Explore genomic #StructuralVariants and #GenomeAlignments in our upcoming #webinar on the new Alignments viewer in the Ensembl Data Platform. It includes a live demo comparing human reference and alternative assemblies. Registration free but essential www.ebi.ac.uk/training/eve... @ensembl.org

PDB Reaches a New Milestone: more than 250,000 entries

PDB Reaches a New Milestone

The PDB now contains more than 250,000 entries

rcsb.org

Unless you use APPRIS principal isoforms, of course, where this can't happen.

Michael Tress@michaeltress.bsky.social · 9mo ago

Finally, this has implications for gene annotation. There is ever more evidence for these isoforms and some have already become the main isoform. If this happens, the N-terminal ext. may get tagged with spurious pathogenic mutations (as in PUS1, TMCO1 and EDNRB) that will be hard to correct.

... OK, I plucked that 17% out of thin air. With @appris.bsky.social principal isoforms as the main representative of each gene, 14.7% of Ensembl reference genes have signal peptides, as do 13.8% of the 14,888 genes we had detected in our proteomics experiments. Exported proteins are in the cell.

Attention developers and researchers working with Ensembl programmatically! Upcoming Ensembl API and data access changes are live on our blog. Read more about the new platform, new services, and migration timelines here: www.ensembl.info/2025/12/02/...

Has APPRIS selected the wrong principal isoform? Send us an email via the adresses on the website and we will investigate.

We have recalculated APPRIS principal isoforms for GENCODE v48 / Ensembl 114 human. GENCODE added loss of function genes to this version and we discovered that this caused the TRIFID scores to change, affecting some of the PRINCIPAL:2 -> PRINCIPAL:5 isoforms. Our fault, we asked for the change!

FlyBase needs your help! Because of recent changes to government funding, the NIH grant that supported FlyBase has been "terminated." flybase.org

We have updated APPRIS to @gencodegenes.bsky.social v48/ @ensembl.bsky.social v114. New principal isoforms for human, mouse, rat, cow and fruit fly ( @flybase.bsky.social ). Principal isoforms for a total of 73 human genes have been updated manually.

Users of APPRIS may have noticed that there is a new category of principal isoforms, Principal:M. These principal isoforms are exclusive to the human gene set and are principal isoforms that have been manually corrected. In GENCODE v47 there were 45 genes with Principal:M isoforms. Here is DCD:

Nice try, Luka. But even in a sham election with the real opposition candidates in jail, that is still some way short. of the 99.86% of supported ClinVar pathogenic mutations that map to APPRIS principal isoforms ...

APPRIS has updated to GENCODE v47 and mouse M36. Danio rerio has also been updated to Ensembl 113. in this release 80.7% of human coding genes have a PRINCIPAL:1 isoform. Also, we have begun the manual curation of principal isoforms. More soon. @gencodegenes.bsky.social @ensembl.bsky.social

APPRIS has new annotations for GENCODE human v46 and mouse M35. The human reference set now has 16,483 genes with Principal:P1 isoforms and mouse has 18,782 genes. Agreement with MANE Select transcripts in human (no MANE Select in mouse) has risen slightly to 17,661 genes. appris.bioinfo.cnio.es#/

APPRIS in use ...

Michael Tress@michaeltress.bsky.social · 2y ago

We have just published a paper in which we take an in depth look at translation of 5' UTR regions. The key takeaways are that it is common and that most translation from 5' UTR is in-frame, so the effect at the protein level is to extend the N-terminal. academic.oup.com/nar/advance-...

APPRIS has annotated principal and alternative isoforms and TRIFID functional isoform scores for Ensembl Bos taurus version e111: appris.bioinfo.cnio.es#/