LikeLTD

LikeLTD models quantitative peak-height information from electropherograms to analyze complex low-template and mixed DNA profiles for forensic interpretation.

Key Features:

• Peak-Height Modeling: Implements a peak-height model that accounts for over-stutter and double-stutter artifacts and uses quantitative peak-height data rather than presence/absence alone.
• Drop-In Analysis: Incorporates a method for handling drop-in events where extraneous alleles appear in a sample.
• Open-Source Implementation: Implemented in open-source R code.
• Model Validation Techniques: Validated using 108 laboratory-generated crime-scene profiles.
• Computational Efficiency: Provides an approximation method to reduce computational complexity when multiple low-level contributors are present but not relevant.
• Linkage Adjustment: Includes a simple approximate adjustment for linkage between loci.

Scientific Applications:

• Forensic Science: Interpretation of crime-scene DNA samples that are low-template or mixtures using quantitative peak-height information.
• Research and Development: Evaluation of benefits and limitations of peak-height modeling in DNA profile analysis.
• Pre-Extraction Replication Assessment: Assessment of pre-extraction replication techniques to improve analysis of low-template samples.

Methodology:

Statistical framework integrating quantitative peak-height data with presence/absence information; explicit modeling of stutter (including over-stutter and double-stutter) and drop-in events; validation on 108 laboratory-generated crime-scene profiles; approximation to reduce computational complexity for multiple low-level contributors; simple approximate linkage adjustment between loci; implemented in open-source R code.