BIN300 W2: Linkage mapping: mapping the SNPs

multipoint linkage analysis

2^(n-1) linkage phases for n loci

• 2,3,4,5, SNPs → 2,4,8,16

need map function

• translates Ɵ to distance

• can estimate interference/coincidence

gene order has to be found by ML

• n!/2 gene orders

Gene orders for 3 SNPs, A, B, C

ABC, ACB, BAC

not CBA, because is ABC read from right to left

multipoint linkage analysis

2^(n-1) linkage phases for n loci

• 2,3,4,5, SNPs → 2,4,8,16

need map function

• translates Ɵ to distance

• can estimate interference/coincidence

gene order has to be found by ML

• n!/2 gene orders

Gene orders for 3 SNPs, A, B, C

ABC, ACB, BAC

not CBA, because is ABC read from right to left

Example 2 loci: ABC

sire genotype

A1B1C1/A2B2C2

Allele inherited by progeny

probability

observed numbersS

likelihood vs coincidence

Alternative gene orders: 3 loci, ABC, ACB, CAB

Alternative gene order: count number of recombinations

order many SNPs → marker map

complicated problem

• n!/2 possible solutions (n=10 yields 2 mill)

• computer cannot test all solutions

• resemblance to travelling sales man problem

• available computer algorithms: simulated annealling

closely linked SNPs

• too little reocmbiantion to find (correct) order

• need sequence data to order closely linked SNPs

  • asummes reference sequence is correct

linkage groups → chromosomes

marker mapping → linkage groups

• linkage group = set of SNPs shown to be linked

• within linkage group SNPs are ordered by mapping

linakge groups form into chromosomes

• if suffiencient dense SNPs to connect all SNPs on chromosome

linkage mapping

• detects SNPs in linkage gropu: ideally: map to chromosomes

• determines order of SNPs: needs sequence data for fine-scale ordering

• computationally difficult problem: many possible orders