6

Due to alternative RNA splicing, it isn't uncommon to ultimately find multiple gene products expressed from one gene in eukaryotes. I'm looking for a reference value for the average number of final gene products expressed per gene for:

  • ... a particular eukaryote (preferably humans).

This can't be too hard to do for one species like humans. I would expect the following formula to provide me with a rough answer. Is this correct?

(number of distinct proteins + numbers of distinct non-translated RNAs) / number of genes

  • ... all eukaryotes as a whole.

This one is slightly more problematic.

LanceLafontaine
  • 6,705
  • 8
  • 36
  • 78

2 Answers2

4

One answer can be found in the UniProt FAQ:

What is the human complete proteome?

In 2008, a draft of the complete human proteome was released from UniProtKB/Swiss-Prot: the approximately 20,000 putative human protein-coding genes were represented by one UniProtKB/Swiss-Prot entry, tagged with the keyword 'Complete proteome'. This UniProtKB/Swiss-Prot complete H. sapiens proteome (manually reviewed) can be considered as complete in the sense that it contains one representative (canonical) sequence for each currently known human gene. Close to 40% of these 20'000 entries contain manually annotated alternative isoforms representing over 15'000 additional sequences ...

http://www.uniprot.org/faq/48

So, we have 20,000 genes and 35,000 products yielding about 1.75 gene products per gene. Alternatively, the 8,000 genes undergoing alternative splicing give about 2.88 proteins per gene.

R Stephan
  • 3,830
  • 18
  • 31
  • I thought it was much higher than this! More in the region of 100's of thousands, because the majority of genes have at least 2 alternative isoforms, usually more... I'll try to find some evidence. If this is the case then I am surprised (also I'm not getting at you - I realize you have cited UniProt, arguably the authority on this!) – Luke Oct 05 '12 at 15:47
  • 1
    The number given here is very debatable... The sequences considered are only manually curated sequences - generally manual curation is way behind the actual data, so it is a bad idea to look at these numbers. For example, see ENSEMBL: http://useast.ensembl.org/Homo_sapiens/Info/StatsTable?redirect=mirror;source=www.ensembl.org , which mentions ~200,000 transcripts. And as I mention below, protein-coding genes are the easy part: ncRNA are much more difficult to count. – Bitwise Oct 05 '12 at 18:26
  • Like the number of genes dwindled in the years after 2000 from 100k to 20k, I expect such unconfirmed numbers to get more close to reality in some years, too. – R Stephan Oct 06 '12 at 05:51
  • @Bitwise, just wanted to point out that the number of transcripts is not necessarily correletaed to the number of different, mature, gene products. A lot of alternative splicing events only affect gene regulation (think spliced UTRs). – terdon Oct 06 '12 at 11:07
1

Well, there are a number of problems. First, the discovery of noncoding RNAs is relatively new and they are difficult to detect, so their number is unknown. In addition, it is unknown how many of them are functional. This has also made it very difficult to define what a gene is and thus made it difficult to count how many genes there are.

Protein coding genes, however, are much easier to deal with, and if you like you can enter any public protein database (at NCBI, for example) and easily query the number of distinct proteins.

Finally, the last problem is that with an organism like human, experimental methods are generally limited by sensitivity (might not detect proteins/RNA at very low concentrations) and the fact that you can't measure all possible cell types/conditions (many gene products will be cell type/condition specific).

Who said life is easy? Somehow things are always complicated in biology...

Bitwise
  • 2,093
  • 13
  • 10