0:00
what happens when you get an allele at a
0:04
locus for one of your markers that has
0:07
an allele frequency of zero if you're
0:11
using pop stands what does it do
0:15
right how do you know its use that in
0:18
pop stats
0:19
oh how do you know then they put the
0:30
number in for you correct there's a
0:34
little M that they print out beside the
0:38
allele you'll see an M that means
0:44
minimum allele frequency and there are
0:48
different ways of calculating it I did a
0:51
bunch of research on this when we were
0:53
trying to get our SOP written for how to
0:56
do this in Denver so I did a little bit
0:59
of research and I found that there's
1:02
actually three three ways of doing it
1:04
that are published first is using two
1:08
pieces of information the Alpha level
1:10
which is your level of confidence right
1:12
95 percent or 5 percent so alpha would
1:15
be 5% if you want to be 99 or 95%
1:18
confident 1% if you wanted to be 99
1:21
percent confident and is the number of
1:24
people in your database okay you know
1:28
that for each group that sometimes
1:30
varies that varies between the groups
1:33
right different number of people so the
1:35
first one from 1992 takes into effect
1:39
accounts for the level of confidence and
1:42
the number of people in the database
1:43
that's the top formula there the next
1:46
one that came out International Journal
1:48
legal medicine in 96 it has a little bit
1:50
more information it uses the Alpha level
1:53
it uses the number of people in the
1:56
database but it also uses C which is the
2:00
number of common alleles at the locus I
2:03
chose that to use that in my when we
2:08
before we were using pop stats I chose
2:10
to use that form of
2:13
the minimum allele frequency because it
2:15
had the most information accounted for
2:18
right and then in NRC two they recommend
2:24
five over 2n okay so you guys are
2:28
critical thinkers what do you think
2:32
between these three what would you do
2:35
and how would you decide well the reason
2:42
I decided on the middle one when I did
2:43
was because it was published the other
2:46
two are as well one is in NRC 96 the
2:49
other ones in a 1992 book by weird I
2:51
chose the middle one because it had more
2:54
information it accounted for more in the
2:57
database right not just the number of
2:59
people but the number of common alleles
3:01
well as it turns out if you look at
3:07
different n sizes numbers of people in
3:10
the database and you look at the NRC to
3:13
5 over 2n you get about at the very top
3:19
0.025 right 150 people vote point oh one
3:26
seven 200 people about 0.01 most of our
3:32
databases that we're using are on the
3:33
order of about 200 people right so you
3:37
can see that 5 over 2 n gives whether
3:40
you have 200 people about 0.01 or 100
3:46
look at the difference between the 3
3:49
methods now over the same number okay
3:53
around 200 people
3:55
Weir's gives point zero zero seven five
3:59
well if you round that up what do you
4:01
get
4:05
what's the point 0 0 75 rounded up 0.01
4:10
right to the next decimal place
4:12
what about Bedok Lee's method that's the
4:16
International Journal legal medicine
4:18
1996 0.01
4:22
what about NRC to 5 over 2 N 1 ok that's
4:28
with the top with a minimum allele
4:31
frequency where alpha is 0.05 and you
4:33
have 8 alleles what about if you have
4:38
double the number of alleles you have 16
4:41
alleles now well what happens for about
4:44
200 people when you get think the
4:53
numbers are slightly different but the
4:54
practical significance isn't right get
4:57
about point or what
