The state of the human coding gene catalogues.

Area of Science:

• Genomics
• Proteomics
• Bioinformatics

Background:

• Previous analysis in 2018 identified significant disagreement among major human proteome repositories (Ensembl/GENCODE, RefSeq, UniProtKB) regarding the coding status of annotated genes.
• The 2018 study suggested over 4000 genes might be incorrectly classified as coding.

Purpose of the Study:

• To re-evaluate the coding status of human genes using updated versions of the three main reference gene sets.
• To assess changes in gene set consensus and identify potentially misclassified coding genes.

Main Methods:

• Comparative analysis of gene annotations from Ensembl/GENCODE, RefSeq, and UniProtKB.
• Identification of discrepancies in coding status across the repositories.
• Evaluation of gene features that may indicate non-coding potential.

Main Results:

• The updated analysis still shows disagreement on the status of 2603 annotated genes (approximately 1 in 8).
• Collaborative efforts have increased consensus, with 249 additional genes now agreed upon as coding.
• Over 2000 genes exhibit features suggesting they may not be coding, with a majority of the 2603 disputed genes falling into this category.

Conclusions:

• Despite improvements, an estimated 3000 genes may be misclassified as coding and warrant re-annotation as non-coding genes, pseudogenes, or cancer antigens.
• Ongoing discrepancies underscore the dynamic nature of gene annotation and the need for continuous refinement of genomic databases.