New leaflet sorter sometimes assigns all lipids to the same leaflet

[EzryStIago]

Old leaflet sorter appears to work here: But the new one fails:

[EzryStIago]

The densities agree. In the second case (new sorter), all the density is in the upper leaflet. POPG.low.avg.dat.txt POPG.upp.avg.dat.txt

[lms464]

@lms464

• [ ] build person config file
• [x] Rerun debug test binary system
• [x] Rerun debug test native membrane

[lms464]

Yeah, there are issues with alg 0. Aligned and Unaligned systems broke the binary system Ezry sent me. Data for check is attached, _data.csv I might have a part of the problem. In the config file

set lipids [list "POPG"] ;# list of all species to bin
set headnames [list "PO4"] ; #lists one headgroup atom/bead name per lipid species 
-> set tailnames [list "C4"] ; #lists one terminal atom/bead name per lipid species 

I think leaving the C4 is resulting in part of the issue Using C4 as the tail type, here is alg 0 compared to alg 1 Exp=Expected count of that lipid Count = what is counted dif = difference between the two upper = upper leaflet lower = lower leaflet diff = different of the two leaflets (+ upp/- low)

   Exp,Count, dif, upper, lower, diff                             Exp,Count, dif, upper, lower, diff
  ,2059,2059,0, 0, 2059, 2059                                  |  ,2059,2059,0, 1032, 1027, -5                                
  ,2059,2059,0, 0, 2059, 2059                                  |  ,2059,2059,0, 1033, 1026, -7                                
  ,2059,2059,0, 0, 2059, 2059                                  |  ,2059,2059,0, 1032, 1027, -5                                
  ,2059,2059,0, 0, 2059, 2059                                  |  ,2059,2059,0, 1033, 1026, -7  

Changing to C4A it appears closer

  ,2059,2059,0, 1032, 1027, -5                                                                                                 |  ,2059,2059,0, 1032, 1027, -5
  ,2059,2059,0, 1034, 1025, -9     difference                                                                          |  ,2059,2059,0, 1033, 1026, -7                                                                                                
  ,2059,2059,0, 1032, 1027, -5                                                                                                 |  ,2059,2059,0, 1032, 1027, -5
  ,2059,2059,0, 1032, 1027, -5                                                                                                 |  ,2059,2059,0, 1032, 1027, -5

While alg 2 is not shown, results have been the same as alg 1 My suggestion is drop the C4 or D4, and choose a acyl chain (A/B) probably which ever is more saturated?