Collaborative Inter-laboratory Validation Exercise of the LRmix Studio software
We get in touch with you to invite you to participate in the eighth collaborative exercise on mixtures, in this new edition, with the perspective of this exercise serving as internal inter-laboratory validation of the LRmix Studio software.
After the last meeting of the GHEP-ISFG in Prague (September 9-10, 2019), in the discussion of the Working Commission on “Autosomal STRs Profiles Mixtures” (GHEPMIX), it was revealed, among other issues, the need for the laboratories to have validated bioinformatics tools for the LR assessment. Among them, the most widely used was the LRmix Studio software. For this reason, we propose the following “Collaborative Interlaboratory Validation Exercise” (GHEPMIX8), whose ultimate purpose is for participating laboratories to have a unique tool for the internal validation of this software, having reference data thanks to the present inter-laboratory study statement on the LRmix Studio version 2.1.5.
There are different ways of carrying out the internal validation of a calculation software such as the LRmixStudio, based on a probabilistic approach to the interpretation of complex mixture profiles, taking into account the “probable” existence of drop-out and/or drop-in. In our opinion, the accuracy of the software must be assessed in order to carry out this internal validation, which will be measured through two parameters: repeatability and reproducibility.
Repeatability is defined as the variation measure in the result when the same test is performed a certain
number of times (at least 3 times) by the same operator, at different times and/or different equipment (e.g., different computers). On the other hand, reproducibility will be the variation measure in the result when the same test is performed by different operators (e.g., by different laboratories).
The objective of this GHEPMIX exercise edition (GHEPMIX08) is to offer to those laboratories that may be interested in the LRmix Studio software validation (in its latest version, v2.1.5), the possibility of using this collaborative exercise as a tool that could contribute to such internal validation, providing a reproducibility measure of the software. The participating laboratory should only internally assess the analysis repeatability. In this way, it is intended to follow the recommendations set by the DNA Commission of the ISFG on the validation of software programs performing biostatistical calculations for forensic genetics applications (Coble et al., 2016; https://doi.org/10.1016/j.fsigen.2016.09.002), in particular, what refers to internal validations (see Recommendations from 10 to 16).
In order to achieve this goal, participants will be provided:
- The exercise development instructions, where the analysis parameters will be clearly established
- (probability of drop-in, Theta correction, ...) that may vary depending on the statement raised.
- 4 problem mixture EPGs (in PDF format) and reference sample profile tables.
- 5 mock cases, simple and clear for the purpose of this exercise, with closed and fixed hypotheses.
- No other hypotheses can be established.
- Data collection form.
- File containing the population database.
- Brief software use guide, for the purpose of the exercise, which in no case will replace the software
- user manual.
The deadlines set for this Collaborative Validation Exercise are:
Registration: until January 10, 2020
Material sending to the participants: January 20, 2020
Results reception: March 27, 2020
Presentation and discussion of exercise results: XXV Meeting of the GHEP-ISFG in Cartagena de Indias
Please, those interested in participating, send an email indicating your interest, as well as the name of your laboratory and contact person, before January 10, 2020 to firstname.lastname@example.org, and CC
to email@example.com, firstname.lastname@example.org and email@example.com.
We hope this exercise is to your liking and we are looking forward your participation.
Working Commission on Mixtures (GHEPMIX)