# Lecture 18: Y Chromosome and mtDNA Markers.

(version 17 March 2008)

This material is copyrighted and MAY NOT be used for commercial purposes

 You are visitor number since 17 March 2008

### Y Chromosome Markers

Often in many cases involving mixtures, DNA from the victim (who is often female) may overwhelm DNA from the perpetrator. Such is often the case for material taken from fingernail scrapings.

In such cases, we can still amplify Y chromosomal markers, which are present in males but not females.

#### Y chromosome markers

Typical test involves looking at 11 STR markers on the Y.

#### Inheritance

The Y chromosome is strictly passed on from father-son. Hence, it tracks paternal lineages

Because of this pattern of inheritance, Y chromosome should track with surnames. Hence, Y markers are now widely used genealogy. For example, see the family tree DNA website

Throughout most of the Y chromosome, there is no recombination.

Hence, markers are inherited as a single group, a haplotype

It is only through mutation that individuals differ.

Hence, a father and all of this sons (and their sons, etc.) all have the same Y haplotype, unless a mutation has occurred.

Mutation rates are on the order of 1/500 per marker

• Prob(father & son identical at 11 markers) = (1-1/500)11 = 0.978

• Prob(two sons identical at 11 markers) = (1-1/500)2*11 = 0.957

#### Match Probabilities

Since the markers are inherited as a single block, the product rule does not apply.

Haplotype frequencies are found by looking up their population frequencies, such as using the YHRD-Y database

Typically the counting method is used.

Suppose there are 4000 samples from Caucasians in the database, and 2 are the haplotype of our sample, freq. = 2/4000

Adjusted counting method: Most haplotypes are rare enough as to not be found in even a modest size database. How do we report the frequencies in such cases?

• Suppose there are n individuals in our database, j of which match our sample

• The suspect is also a new individual, making n+1 individuals, j+1 of which match

• To be extremely conservative, we suppose that the crime sample came from someone OTHER than the suspect, this adds an additional match and an additional individual to the database

• Thus, Pr(Match )= (j+2)/(n+2)

• A special case: No match in the database, Pr(Match )= 2/(n+2)

Typical match probability is around 1/1000 to 1/5000. This is entirely a function of the size of the databases. As the database exceed 10,000 (and higher) the match probabilities will continue to go down.

#### Computing joint Y-autosomal match probabilities

Use product rule: Autosomal match probability * Y match probability.

### Mitochondrial DNA (mtDNA) Markers

With a very degraded sample and/or tissues lacking any nuclear DNA (such as air shafts, bones, and teeth), autsomal markers may either not be present or so degraded as to be useless. In such cases, mitochondrial DNA (mtDNA for short) can prove a powerful tool.

#### The Mitochondria

A typical human cell contains several thousand mitochondira, where each mitochondira contains tens to thousands of (identical) mtDNA molecules.

The mitochondiron is the "powerhouse" of the cell, generating energy. Hence, they are found in all cells, even those specialized cells that have lost their nuclear DNA (and hence any autosomal markers).

#### mtDNA

Thousands of mtDNA molecular per cell, each of which is around 17,000 bases long and circular. Its this high copy number, and the fact that ALL cells have many copies of mtDNA, even when they do not contain a nucleus, that makes mtDNA a useful forensic tool in some settings.

#### mtDNA markers

Here an entire short region, around 780 bases, is sequenced (from the mtDNA control region). It is then compared to a standard (the Cambridge reference sequence, the first published human mtDNA sequence). Deviations from the CRS are reported.

Two random individuals from within any particular population are expected to differ on average by 11 +/- 5 bases in the control region.

One novel issue with mtDNA is heteroplasmy, where not all of the copies (in the thousands of copies per cell) are exactly alike. This occurs at detachable levels in a small fraction (less than 10 percent) of individuals, and arises because of mutations. Hence, an mtDNA "haplotype" may actually be a couple of sequences that are identical except for being off at one (or two) bases.

There is no recombination on the mtDNA

#### Inheritance

mtDNA is strictly passed on from the mother to all her offspring. Hence, it tracks maternal lineages

#### Match Probabilities

Use modified counting rule, again need haplotype frequencies for data bases.

Currently, match probabilities are around 1/100 to 1/1000.

#### Computing joint mtDNA-autosomal match probabilities

Use product rule: Autosomal match probability * mtDNA match probability.