This page lists references used for implementing specific functionality into Pylake. Please see below for more information on these topics.
The passive force calibration method was based on a number of publications by the Flyvbjerg group :cite:`berg2004power,tolic2004matlab,hansen2006tweezercalib,berg2006power`.
Several polymer models from literature were included. These include Odijk's extensible worm-like chain model :cite:`odijk1995stiff,wang1997stretching`, the Freely Jointed Chain model :cite:`smith1996overstretching,wang1997stretching`, the Marko-Siggia interpolation model :cite:`marko1995stretching` and the twistable worm-like chain model :cite:`broekmans2016dna,gross2011quantifying`.
We used :cite:`broekmans2016dna` as reference material throughout the implementation of the F,d fitting routines, both as a starting point for the model implementations as well as for the procedure on how to reliably fit the twistable worm-like chain model (via its inverted form). For the twistable worm-like chain, we perform model inversion via interpolation rather than point-wise inversion through optimization analogously to :cite:`broekmans2016dna`. Parameter estimation is performed via Maximum Likelihood Estimation :cite:`maiwald2016driving,raue2009structural`. We also implemented several asymptotic information criteria for ranking model fits :cite:`cavanaugh1997unifying`.
The line algorithm was based on sections 1, 2 and 3 from :cite:`steger1998unbiased` to find line centers based on local geometric considerations. The greedy algorithm was based on two papers. It initially detects feature points based on :cite:`sbalzarini2005feature`, followed by line tracing inspired by :cite:`mangeol2016kymographclear`.
.. bibliography:: refs.bib :style: alpha