In genome-wide association studies (GWAS), what is the difference between imputing and typing? Are they the same?
'Imputing' is an imputation. It is a statistical inferring from existing genotype data using haplotype estimation. It is useful because different genotyping arrays genotype a slightly different set of SNVs and they can't detect all variants as whole genome sequencing is still pricey.
No, they are different altogether.
Typing in GWAS is genotyping individuals from a selected population to know say the genetic variation of SNP's (single nucleotide polymorphisms) among them via chips like Illumina Core-Exome etc.
Imputing is the statistical process of imputation of SNP's of an individual from a reference source like the HapMap reference panel. The process is a bit complex and typical tools like IMPUTE2/BEAGLE4/MINIMAC2 rely on HMM (Hidden Markov Models) to fill in the SNP's.
So, in essence within a GWAS study first, we genotype individuals (as it's cheaper than sequencing all SNP's) and get 300-500k SNP's via the chips as above, then impute the remaining variants using an established Human Genome Reference panel using the above software.
An excellent resource for understanding imputation is Yun Li's 2009 paper.