[simuPOP-list] Forward simulation of DNA recombination from 16 defined individual ancestor

[Neil M. <nei...@gm...>]

Because of the simplicity of my intended modeling, I fear that using a robust program like simuPOP might be like swatting a fly with a hammer.

My hope is that by using the program, I can gradually increase the complexity of my models, and so want to use simuPOP at the outset.

I am working on a genealogy matrix derived from downstream recombination of DNA segments.  I intend to ignore any consideration of mutations or variations in recombination rates across the chromosomal board.  

By example, I will start with a population of 16 defined ancestors, designating their two respective paired maternal and paternal chromosome sequences something like ‘AAAAAAAAAAAA’  — ‘aaaaaaaaaaaa’  through ‘PPPPPPPPPPPP’ — ‘pppppppppppp' (of course much longer)

The first generation offspring of A and B, might have recombinant sequences that look like  'AAAaaaAAaaaaa’ and ‘bbbbbbbbbBBBBb’ randomly recombined without attempting variations across the chromosomal map.  I would like to simulate recombination rates per million and be able to adjust this in different simulations.

Future generations will each contribute to a cascade of descendents, each of which will have a decreasing component of their respective direct original 16 ancestors.  External DNA contributions, increasing with each generation, can simply be designated as ‘Z’ as they will not be considered in the intended analysis.

My goal is to model the frequency of cross-matched segments in descendant generations, and compare it with known human DNA segment match data. 

My first line goal is to develop multiple instances of grouped match triangulations datasets, and analyze them with respect to known and defined relationships among the entire descendant population.

Is this something that seems appropriate for this software, and could you gently poke me toward the correct way to achieve this ?

Thank you in advance for your consideration and help

Neil Millikin
(retired scientist with an interest in applying computer science to genetic genealogical analysis)

I am quite proficient with Python data analysis, but freely admit to getting lost when reading the source code.