Estimating error rates for single molecule protein sequencing experiments

Area of Science:

• Biotechnology
• Computational Biology
• Analytical Chemistry

Background:

• Single molecule protein sequencing (SMPS) technologies are emerging, but their practical use depends on accurate error rate estimation.
• Fluorosequencing, a type of SMPS, generates data that requires specialized analysis methods to account for inherent errors.
• Existing methods may not fully capture the complexities of SMPS error profiles.

Purpose of the Study:

• To develop and evaluate novel parameter estimation methods for analyzing single molecule protein sequencing (SMPS) data from fluorosequencing.
• To accurately quantify error rates associated with SMPS, including missed cleavages, dye loss, peptide detachment, and N-terminal blocking.
• To compare the performance of a Hidden Markov Model (HMM) based approach with a hybrid optimization method (DIRECT and Powell's).

Main Methods:

• Development of a Hidden Markov Model (HMM) approach, extending 'whatprot', utilizing a modified Baum-Welch algorithm for parameter estimation.
• Implementation of a second method combining DIRECT and Powell's optimization techniques to minimize root mean squared error (RMSE) between simulated and real SMPS data.
• Validation of both methods using simulated datasets and experimental fluorosequencing data, including comparisons against controlled experimental perturbations.

Main Results:

• The HMM-based approach demonstrated high accuracy on simulated data and provided reasonable parameterizations for experimental datasets after incorporating N-terminal blocking and pre-processing.
• The hybrid DIRECT and Powell's method was developed to reduce RMSE between simulations and experimental data.
• Comparison showed the Baum-Welch based HMM approach outperformed the hybrid method by most criteria, though both yielded similar error rate estimates for experimental SMPS data.

Conclusions:

• The developed HMM-based parameter estimation method offers a principled way to analyze fluorosequencing data and estimate critical error parameters.
• Both evaluated methods provide valuable insights into SMPS error rates, with the HMM approach showing superior performance in most aspects.
• Accurate error rate estimation is essential for advancing the practical application and reliability of single molecule protein sequencing technologies.