NEW YORK (GenomeWeb News) – The rate at which de novo mutations pop up in the germline is much more variable than previously appreciated, according to a study in Nature Genetics online yesterday.
Members of the 1,000 Genomes Project used probabilistic algorithms, targeted re-sequencing, and haplotype phasing to assess de novo mutation rates in two parent-child trios — one of European descent and another of Yoruban ancestry — whose genomes had been sequenced for the pilot phase of the international effort. Their results suggest each parent passes on an average of thirty germline mutations to their children, though the proportion of mutations coming from each parent varied from one parent to the next and also between families.
"The variation in mutation rates that we observed is of potential clinical importance, as it suggests that the risk of misdiagnosing a [de novo mutation] as being pathogenic could vary substantially between individuals," co-corresponding author Philip Awadalla, a medical and population genomics researcher at the University of Montreal, and his co-authors wrote.
Researchers have long speculated that males might pass on more germline mutations than females, since far more cell divisions are needed for sperm than egg production, upping the chances that glitches will occur when DNA is copied and partitioned into these sex cells.
Even so, germline mutation rates remain murky, the study's authors explained, since past estimates of germline mutation rates have relied on indirect measurements based on substitution rates between related species, such as humans and chimpanzees, or from mutation rates extrapolated from disease studies.
For the current study, Awadalla and his colleagues set out to directly determine the rate of de novo mutation in male and female germlines using genome sequence data from the 1,000 Genomes Project pilot.
The genome sequences had been generated from lymphoblastoid cell lines created using samples from two families — a child and both parents from the Yoruban population and another parent-child trio of European descent.
To track down germline de novo mutations, members of the team developed probabilistic algorithms for finding these new mutations in the genomes, each sequenced to an average of nearly 23 times coverage using a combination of Illumina, Roche 454, and SOLiD sequencing platforms.
These computational tools "specifically assign a probability to a call being a de novo mutation," Awadalla told GenomeWeb Daily News. "That allowed us to draw a circle around all the potential de novo mutations that there were from two families that were sequenced in the 1,000 Genomes Project at high depth."
After sifting through thousands of potential mutations, the team verified genuine germline mutations by using Illumina and SOLiD sequencing platforms to sequence selected sites in DNA from lymphoblastic cell lines and from either blood samples, for trio members from one family, or from additional family members from a third generation in the case of the other family.
Overall, researchers tracked down and validated 49 germline de novo mutations in the European trio and 35 in the Yoruban trio, along with nearly 1,600 non-germline mutations representing somatic mutations or mutations that cropped up in cell lines generated from the individuals' samples.
By drawing on the longer reads that had been generated using the Roche 454 platform, in combination with SNP data, the team was then able to phase mutations based on linked alleles, helping them determine which de novo mutations came from each parent in the trios.
In one of the families, most germline de novo mutations — 92 percent of those detected — appear to have been inherited from the father. The other family showed a very different pattern, though, with 64 percent of germline de novo mutations coming from the child's mother and just over a third from the father.
Although more research is needed to determine how much of the difference in de novo mutation rate is a consequence of inter-individual variation in general and how much stems from variations in the de novo mutation rates involving different gametes from the same individual, the team noted, the findings hint that a wide range of germline mutation rates exist in humans.
"With only a single offspring per family, we could not distinguish between these two alternatives," they wrote, "but either would give rise to substantial variation in the number of [de novo mutations] between offspring of different families."
In the future, the team plans to look at more families sequenced through the 1,000 Genomes Project and other studies, Awadalla said. They are also keen to assess mutation patterns in multiple tissue types to get a better sense of the somatic mutation rate in humans. Because it appears that inter-individual variability exists in the fidelity with which mistakes are repaired during DNA replication, Awadalla noted, there is likely more to be learned about the genetics and heritability underlying human mutation rates as well.