withoutR/cpp/README.md

In bioarch-sjh/bacollite: Analysis of MALDI and LC-MS/MS spectral data using sequence data (mainly for collagen)

Strategy

This has been a bumpy ride, but the way we call cpp functions (starting from the cpp end) is as follows:

• Develop pure cpp code in the src/ directory, but use symbolic links to test in withoutR\cpp
• Write a cpp test file, in withoutR/cpp
• Write a cpp wrapper that uses R data structures and converts between
• Create R using devtools::document() and or devtools::build()

How R_iso_seq works

• calls pepFromSequence, which subtracts the ASCII value for A (65) from the letter to give an integer representation of the amino acid. These codes are stored in the NUMSEQ variable, and the whole structure is passed as the first argument to R_iso
• calls allocDistArray to allocate the distributions - this was needed when >1 peptides were being processed at once..
• calls R_iso to get the isotope distributions

How R_iso works

This should form the basis of the R_iso_atom function that we are trying to write

• numpep = 1, numiso = 5, otherwise it won't work
• load amino acid table AMINODATA
• create the following variables:
  • peptide - a double of size NUMELEMENTS
  • element - This is an array of ISOTAB structs. This is populated using the R_loadIsotopeTable function. The original iso function parsed >1 peptide, here numpep always == 1. Here the word "element" means the component of the total isotopic mass (I think).
  • letter - letter[i].element[k] = AMINODATA[i][k] where 26 letters i and five elements k are used - showing the composition of elements in each of the 26 amino acid codes. - Enter a for loop which only iterates once because of the value of numpep
• accumulate the number of elements in the appropriate slot of the peptide variable

bioarch-sjh/bacollite documentation built on Oct. 7, 2022, 3:34 p.m.