Contents
other newsletters

DNA·VIEW NEWSLETTER #18

February 20, 2003

New capabilities in DNA•VIEW

STR Mutation Model

Standard paternity case calculations incorporate a realistic step-wise mutation model, beginning with Version 25.46. The essence of the model, which applies (optionally) to STR loci, is that 1-step paternal mutations occur at the rate specified by the per-locus "mutation rate" parameter, 2-step etc. or maternal mutations less frequent by specified ratios. There are a few other details, for which please see http://dna-view.com/mufeatur.htm. This feature is the realization of the point of view that I described several years ago at http://dna-view.com/mudisc.htm.

SNP's

We are beginning to see success identifying World Trade Center victims with the experimental SNP technology.

In theory no enhancements were needed to DNA·VIEW to cater to SNP's; they could be treated as a 4-allele or 2-allele STR system. However, several changes make dealing with SNP systems more convenient.

In addition, I've made some computations to investigate the impact of using loci that are occasionally close together. First, it is surprising how little linkage association there appears to be. From the data I have examined there is no sign of linkage association between SNP markers 1 centimorgan apart within the major US population groups, and association is not obvious even at a 0.1 cM. Second, although linkage within a family (even when there is no linkage association in the population) is a theoretical problem for some kinship problems, the practical effect of ignoring linkage seems to be slight.

Brilliant Simulation Feature

Geraldo wants to know if DNA testing can confirm family legend that his mother was an illegitimate child of the wealthy railroad baron, Midas Croesus. If he can get cooperation from two of his putative cousins, will he be able to prove or disprove the relationship?

Automatic Kinship has a brilliant new option, Simulate, that answers questions like these. that answers questions like this. The idea got some impetus several years ago when I was given a paper to referee called "Prospective Kinship Analysis." The authors tried to analyze and predict the likely benefit of DNA typing to resolve some particular inheritance dispute. The concept was good but the execution was not. I suggested rewritting but I never saw it again. I felt constrained from trying to write on the subject myself after that, but saw no reason not to write a computer program.

The general problem that Simulate tackles is this: Given a collection of people and conflicting hypotheses as to how they may be related, how helpful is future DNA testing (or further DNA testing, as the case may be) likely to be in resolving the conflict?

The method of solution is to assign genotypes at random simulating nature's laws of genetics, compute the consequent likelihood ratio, and do this many times. The result of running the Simulate program is thus not a single number nor a yes/no answer, but a large collection of alternate possible numbers, whose range and distribution mimic the range of expectable results from DNA typing. From this output you can judge whether a decisive result is certain, likely, or unlikely.

See http://dna-view.com/simulate.htm.

DNA profile simulation

A new feature of DNA odds – useful for amusement or demonstrations - lets you generate random DNA profiles. Enter # for an allele and the program will substitute an allele randomly selected based on the frequencies of the relevant database.

Population analysis tools & enhancements

Allele report

An ordered table of all alleles for a locus and their frequency over all (selected) databases is a helpful addition to the Allele report
Rare allele report. locus: TPOX 2p25-p24 STR
All alleles (*=beep when typed in):
6
7
8
9
0.02
0.0071
0.45
0.13
10
11
12
13
0.066
0.27
0.055
0.0013
*14
*15
1/2900
1/23000

Asterisks flag alleles that are not part of the allele resolution filter (ARF) if there is an ARF, i.e. those that cause a warning beep when typed in.

Computer lore

Host Access Error

In the process of transmitting files from one computer to another there are various ways that the "Read only" attribute can get turned on. For example, copying a file from a CD usually has this result. If DNA•VIEW or PATER then tries to write to that file, the result is the error message HOST ACCESS ERROR.

You can remove the read-only status and fix the problem by right-clicking on the file name in Windows and changing the "properties".

Windows XP startup glitch

Starting DNA•VIEW or PATER under Windows XP is likely to produce a pop-up window with the message:
16 bit MS-DOS Error
The system cannot open COM1 port requested by the application. Click 'Close' to terminate the application
Click 'Ignore' and the program will start ok. To avoid the message altogether, edit two files:
  1. In dnaview\dnaview.bat, remove the E2 or E256 at the end of the line (near the end of the file) that includes APLPDX.
  2. In dnaview\aplii\config.apl, remove any line that begins Env=

Printing from Windows XP

Printing from DNA•VIEW or PATER when running under XP is tricky because they print to LPT1 (or LPT2) and Windows does not allow user access to those devices. Moreover, unlike Windows 98, Windows XP does not have a "capture printer port" option on the printer properties window.

The effect can be achieved, however, with the NET USE command

Step 1 – determine the official name of the printer
Control Panel, Printers and Faxes –
Right-click on the printer, Properties
Click the "Ports" tab. Notice the list of ports. The first column is "Port". Make it wide enough to view the full name. For example, I see
Port
\\ODYY\ODYY 970 USB
Step 2 –Invoke the NET USE command
net use LPT1 "\\ODYY\ODYY 970 USB" Persistent=yes
Note the use of quotation marks around the resource name. They permit designating a printer whose name has spaces in it. Persistent=yes means persisting after the computer is shut down and restarted.
Step 3 – test
As a quick test, type, still at a DOS prompt:
copy testfile lpt1
where "testfile" is some small test print file such as \dnaview\examples\elvis.kin.