Your genome carries up to 11 broken bitter-taste genes—dead copies that once built working receptors and now do nothing at all.

Alongside 25 working bitter-taste genes, the human genome carries somewhere between 7 and 11 pseudogenes—broken copies of the same family that no longer build functional receptors.

Your tongue detects bitterness through a family of receptor proteins built from 25 working genes called TAS2Rs. But scattered through the same stretch of your genome are the wreckage of others: broken copies of these very genes, mutated past the point of function, that once built receptors and now build nothing. Depending on how you count, there are between 7 and 11 of them. They are still there, faithfully copied into every cell you carry, doing absolutely nothing.

These are pseudogenes—the genomic equivalent of a rusted machine left in place after the factory moved on. A pseudogene starts as an ordinary, working gene, then accumulates a mutation that jams the assembly line: a premature stop codon that cuts the protein short, a shift in the reading frame that scrambles everything downstream, a deletion that removes a critical piece. The DNA sequence remains, recognizably related to its functional cousins, but the instructions no longer produce a receptor that can sit in a taste cell and register a bitter molecule.

Why does the count depend on who's counting?

The disagreement between 7 and 11 isn't sloppiness—it's annotation. Deciding whether a stretch of DNA is a true pseudogene or something else requires judgment calls about how broken a gene has to be before it stops counting as a gene at all.

Sanghyeon Oh and colleagues, writing in *BMC Evolutionary Biology* in 2014, put the number at 11, describing "25 functional TAS2R loci" plus "11 TAS2R pseudogenes." That higher figure traces back to earlier work by Yoshiro Go and colleagues in 2005. But Davide Risso and colleagues, in a 2022 survey of the whole bitter-receptor family published in *Frontiers in Genetics*, took a more conservative view: "The complete TAS2R repertoire in humans includes 25 annotated functional loci and seven pseudogenes." The same Risso group had earlier used the figure of 11 in a 2016 *Scientific Reports* paper, which shows how these counts shift as reference genomes and annotation standards are refined.

What everyone agrees on is the anchor: 25 functional genes. That number is corroborated across all three papers. It's the graveyard that's harder to census.

How we know these are dead, not just different

The tell is in the sequence itself. A functional TAS2R gene reads as an unbroken instruction for a receptor protein—a single open stretch that a cell can translate from start to stop. A pseudogene carries the disqualifying flaw right in the letters: the stop codon that appears too early, the missing base that throws off the reading frame. Researchers don't have to test the protein in a dish; the corruption is legible in the DNA, and it's shared across humans everywhere, which tells you the break happened long ago and got passed down.

That's also what makes these fragments interesting rather than merely dead. Oh and colleagues titled their 2014 paper around the idea that variation in these pseudogenes provides "evidence for a dynamic role in human evolution." A broken bitter receptor is a record of a receptor our ancestors once had and no longer needed—or could no longer afford. Bitter taste evolved largely as a poison-detection system, a warning that a plant might be toxic. When a population's diet changes, a receptor tuned to a compound it never encounters anymore stops being maintained, and mutations that would have been fatal to function slip through unpunished.

Here's the seam still open to argument. It's easy to read a pseudogene as pure junk—a fossil, inert. But a growing body of work across the genome suggests some pseudogenes still get transcribed into RNA and may quietly influence how their functional relatives behave. Whether any of the human TAS2R pseudogenes do anything at all—or are truly the silent ruins they appear to be—is a question the sequence alone can't settle. The gravestones are counted. What, if anything, still stirs beneath them is not.