Hi there,
What methods are typically used to determine zygosity with the SAM files located on the 1000 genomes FTP?
I've noticed that in the case of very low coverage (for example, 1x), samtool's pileup maq consensus column will state that the consensus basepair is whatever it has a read for.
But in the case of 1x coverage, how can we be sure that this individual is truly homozygous for some allele at a position? Wouldn't this require higher coverage? My understanding is that the read is, with a 50/50 chance, either from the paternal or maternal chromosome.
Could someone post a link explaining how reads are spread between the maternal and paternal chromosome?
Am I correct in thinking that the sequence reads are evenly split between maternal/paternal chromosomes (as number of reads-->infinity)? If this is the case, I figure that using a binomial, with p=0.5 (assumption), and a desired phred consensus score of 20, we would need at least 7 reads with phred score >=20 in order to know we have read both the paternal and maternal chromosome, and be able to say that an individual is either heterozygous or homozygous for some allele with a phred-probability of 20 or greater, as ((1-0.5)(0.5))^7<0.01.
With less than 7 reads, I would only be able to say "this individual has at least one of this allele at this position", not knowing if that individual is homozygous for that allele, or heterozygous, correct?
Any help would be greatly appreciated.
-- Alex
What methods are typically used to determine zygosity with the SAM files located on the 1000 genomes FTP?
I've noticed that in the case of very low coverage (for example, 1x), samtool's pileup maq consensus column will state that the consensus basepair is whatever it has a read for.
But in the case of 1x coverage, how can we be sure that this individual is truly homozygous for some allele at a position? Wouldn't this require higher coverage? My understanding is that the read is, with a 50/50 chance, either from the paternal or maternal chromosome.
Could someone post a link explaining how reads are spread between the maternal and paternal chromosome?
Am I correct in thinking that the sequence reads are evenly split between maternal/paternal chromosomes (as number of reads-->infinity)? If this is the case, I figure that using a binomial, with p=0.5 (assumption), and a desired phred consensus score of 20, we would need at least 7 reads with phred score >=20 in order to know we have read both the paternal and maternal chromosome, and be able to say that an individual is either heterozygous or homozygous for some allele with a phred-probability of 20 or greater, as ((1-0.5)(0.5))^7<0.01.
With less than 7 reads, I would only be able to say "this individual has at least one of this allele at this position", not knowing if that individual is homozygous for that allele, or heterozygous, correct?
Any help would be greatly appreciated.
-- Alex
