Contents
other newsletters

DNA·VIEW NEWSLETTER #17

March 19, 2002

New capabilities in DNA·VIEW

DNA·VIEW is being used to help with identification of bodies from the World Trade Center collapse. Some of the enhancements added for that work are also useful for the general user, and are among the features discussed below.

Automatic Kinship

Automatic Kinship is a new command, a modified version of Paternity Case particularly designed for kinship work. The list of possible roles – i.e. M for mother, etc. – is extended for mnemonic convenience.

New Immigration/Kinship options/features

The new option Estimate likely relationships under immigration/kinship performs calculations comparing every pair of individuals in the case, and exhibits a likely relationships table such as:
M N F U
V pi=3000
si=30
pi=200
si=10000
pi=20e6
si=1e6
pi=30
si=200
M pi=70000
si=700
pi=400000
si=30000
N pi=50000
si=4000
pi=40000
si=1e6
F pi=1000
si=30

The numbers in each cell evaluate the corresponding pair of people as potential parent-children (pi), as potential siblings (si) – which can of course also be considered as a surrogate statistic for less immediate relationships – and, if the genotypes are consistent, as identical (id). A number is omitted if it is very small.

The above example indicates for example that VF appear to be related as parent and child, and that N may be a child of M and a sibling of V.

Select Add relationship estimates to report to include the table in the report.

The Next race option makes it easy to cycle among the races listed in the race list, if there is more than one.

If you specify a prior probability, it will be used to convert the likelihood ratio to a posterior probability that the principal hypothesis is correct assuming that the only alternative is the alternative hypothesis.

Kinship cannot properly take into account a possible mutation, but for a rough indication you can turn on the consider mutation? toggle.

The display of symbolic genotypes has been enhanced for easier visual assimilation –

D3VWAFGA AmelD8D21D18D5
V
M
N
F
U
 qr
pq
 qr
 qr
pq
V
M
N
F
U
 q
 q
pq
pq
pq
V
M
N
F
U
pq
 qr
p r
p
p r
V
M
N
F
U
xy
x
x
xy
x
V
M
N
F
U
 qr
pq
 qr
  r
 qr
V
M
N
F
U
p
pq
pq
p
pq
V
M
N
F
U
 q s
pq
 qr
  rs
p r
V
M
N
F
U
pq
 q
pq
p
pq
V
M
N
F
U

Crime Case – synthesis of profiles

The Crime Case command, which makes a likelihood ratio computation comparing two specified hypotheses for a mixture, has a nice method to synthesize mixtures without having to type them in. Instead, just type in the profile for each person, because now you can specify the mixture as a composite of several people.

If the mixture has extra alleles in addition to the composite, create an extra, pseudo-person consisting of those extra alleles. The mixture can then be designated as the combination of the pseudo-person and appropriate real profiles. Similarly, when the mixture consists of alleles representing nearly a continuum of signal intensities, create several pseudo-people corresponding to various ranges of peak height. It is then easy to try alternate computations using different assumptions as to which peaks are relevant.

Check for mislabeled samples

The command Screen disaster matches, although written to compare victim and reference profile lists, can also be useful to locate a profile that was mis-coded into the wrong paternity case. Just screen a membrane against itself. If any man closely resembles a child from a different case on the same membrane, the duo will show up on the report.

Y-haplotypes

The new command Y-haplotype odds counts the number of occurrences of a given Y-haplotype in a reference database.

Population analysis tools & enhancements

Validation of imported databases

The Import genotypes for database option of the command Import/Export (in the Input/Output menu) performs several validity checks: duplicate profiles, nearly duplicate profiles, duplicate sample numbers.

Allele report – rare and common alleles

The allele report command, in the "Research" menu, lists all occurrences of sufficiently rare alleles for all (or a selected subset) of the available databases for one locus at a time. It also enumerates the popular and infrequent alleles:

D3S1358. Rare alleles (frequency at most 1/1000 among all 12 databases):  9 11 152 153 21
common alleles:  15-31% 16-29% 17-21% 18-9% 14-8%
		other alleles: 12-0.1% 13-0.5% 19-0.7% 20-0.1%
Allele 9 observed 1/5276
	9 	1/390 	ABI Black 	195	97/08/01 9am
Allele 11 observed 4/5276 = 1/1300, in 3/12 databases.
	11 	1/390 	ABI Black  	195	97/08/01  9am
	11 	1/400 	ABI Caucasian 	200	97/08/01  9am
	11 	2/420 	FSC-Jul99 Black	210	01/07/06 11am
etc.

Database statistics

An enhanced collection of summary statistics for a specified collection of databases are compiled by the option Analyze database statistics of the Database command. The new statistics include the expected stain matching performance and expected paternity index (with or without the mother) for a battery of tests.

Cumulative mean statistics for 9 Malaysian Chinese databases –
     D3S1358 VWA FGA D5S818 D13S317 D7S820 TH01 TPOX CSF1PO
     Matching chance=1/1.9e9  A=99.9%
     PI=6190(=2.64/locus)  mPI=291(1.88/locus)

Cumulative mean statistics for 9 Malaysian Malay databases –
     D3S1358 VWA FGA D5S818 D13S317 D7S820 TH01 TPOX CSF1PO
     Matching chance=1/3.2e9  A=99.92%
     PI=8000(=2.71/locus)  mPI=357(1.92/locus)

Null alleles

Null alleles are now entered as part of each database, rather than being defined as a fixed proportion for each locus. Use Type in (or Edit) a database and type n as the allele name for nulls.

Miscellaneous

Other new features

Locus ordering menu

Whenever the locus menu is presented, you may hit home to obtain a new pop-up menu of popular locus batteries and multiplexes in their standard orders. Select from this list to customize your locus list with move the selected battery, in order, at the top.

Menu selection by context

Users who like to type have traditionally (beginning at least 1988 for DNAVIEW; later for Windows) been able to type the leading characters of the desired menu item. A better method, implemented in recent DNAVIEW versions, allows designation of a menu item by typing a string of characters that occurs anywhere within the desired line. The effect of this revolutionary method is startling at first, but after a small amount of practice it becomes natural, and is easier and more flexible than the old way, especially if you read the documentation to learn a few details and useful tricks.

Clear, delete, or rename a membrane

From the menu of membranes (worklists), move the red bar to a membrane of interest and hit del. A pop-up menu appears. Options include expunging the membrane completely, deleting the reads associated with it, clearing the roster, and changing the name.

Selection of Options

The Options command has improved presentation so that for several of the options that are toggles, you no longer need to select the option to see its state; the current setting displays directly in the Options menu.