Paper describing ReSOLVIt published in FSI:Genetics

We modify our full in silico DNA pipeline to focus on resolving signal-to-noise within the single-copy regime.

The paper describes system design and ways in which the system can be implemented into forensic DNA validation and optimization.

Paper Highlights

•An in silico DNA system is parameterized with the laboratories own experimental data, resulting in predictions of optimal laboratory settings.
•Noise and allele peak height distributions from a single copy of DNA are used to assess signal-to-noise resolution.
•Optimal signal detection thresholds, or analytical thresholds, for casework are obtained, if necessary.
•We demonstrate that metrics of signal quality for simulated and experimental data are consistent.
•This is a systematic method for evaluating EPG quality and is a critical step towards standardizing the post-PCR process.

Drs. Lun & Grgicak speak to the Legal Aid Society

On October 26th, 2017 we will speak at a conference hosted by the Legal Aid Society. See our calendar of events for location and other details.

Dr. Desmond Lun will be speaking on our research pertaining to the development of computational methods in forensic science.

Dr. Catherine Grgicak speaks about signal quality and effects on continuous Likelihood Ratios.

PROVEDIt: Release of Computational Tools and FSA/HID Files at 2016 ISHI


Producing an empirical data set large enough to efficiently compare, contrast, and validate forensic DNA computational systems is costly, labor-intensive, and requires the amplification of many samples that may not be readily available.  In an effort to provide continued support to the forensic science community and to foster growth in both forensic research and operations, we announce the PROVEDIt initiative:  Project Research Openness for Validation with Empirical Data.

First announced at the 2016 International Symposium on Human Identification, PROVEDIt comprises 25,000 .fsa and .hid profiles as well as a suite of analysis, interpretation, and in silico software systems/procedures and models developed in a variety of environments by a multi-disciplinary, inter- institutional team.