cheminfo / chromatography Goto Github PK

Currently the filter is an inplace modification.

Would be nice to also have as an option the possibility to get a copy of it.

Agilent chemstation can export data as 'AIA' format that is inf act a NetCDF

for MS it might be interesting to compute the SPLASH hash (https://splash.fiehnlab.ucdavis.edu/, https://www.nature.com/articles/nbt.3689) which might be neat to search databases. I can look into it.

Seems that if the data are in min in the jcamp it is not taken into account.

Today for series 2D we have

    [
      [
        [
          127.5, 83.7, 75.9, 63.9,
           55.9, 43.9, 39.9,   38,
           36.8,   36, 34.8, 31.9,
             29,   28,   18,   17,
           15.9, 13.9
        ],
        [
            58,  33, 227,   31,   38,
           948,  73, 451,   42, 1017,
           258, 905,  42, 1908, 5593,
          1444, 147,  60
        ]
      ],

An array (the time) of array (2 values, mass, intensity) of array (the mass values and the intensity values).

I would like to change this to

    [
      { 
       x :  [
          127.5, 83.7, 75.9, 63.9,
           55.9, 43.9, 39.9,   38,
           36.8,   36, 34.8, 31.9,
             29,   28,   18,   17,
           15.9, 13.9
        ],
       y : [
            58,  33, 227,   31,   38,
           948,  73, 451,   42, 1017,
           258, 905,  42, 1908, 5593,
          1444, 147,  60
        ],
      },

This will allow me to add a third parameter (third dimension) that is in this case the msms. We have indeed mass spectra of GC / MS / MS

This would be of course a major release and will also require major release of jcampconverter and mzdata.

@targos Is this ok for you ?
@josoriom This should solve your problem as well.

For example, I don't think that has any sense to do the calcultateTic always in place

applyLockMass('mf', options);

• Calculate from mf the monoisotopic mass

var CC = require('chemcalc');
var info = CC.analyseMF(mf);
var em = info.em;

• Find the closest mass value to this em in the lock mass spectrum
• Calculate the difference between theory and experimental
• Remove all the spectra concerning the lock mass
• Apply the difference to each experimental spectrum (the difference change from one spectrum to another)

I updated netcdf-gcms to use the new netcdfjs that uses the new iobuffer that use utf8 v3 instead of v2.

[email protected] /Users/lpatiny/git/cheminfo/chromatography
└─┬ [email protected]
  └─┬ [email protected]
    └─┬ [email protected]
      └── [email protected] 

There is now a problem with the build

cheminfo-build --entry src/index.js --root Chromatography
.
Error: 'decode' is not exported by node_modules/utf8/utf8.js

@targos Can you help me with this problem ?

Get inspired by:

https://github.com/NPellet/visualizer/blob/master/src/lib/chemistry/gc-generator.js

But without dealing with annotations and labels

There could be an option to select the range

• add a folder 'util' and other ?
• use cosine from mljs ?

I think we may assume that the time is a sorted array in increasing value

The algorithm could be improved in order to take this into account and make a dichotomic approach. Pretty sure that there is a library that must do this

Please check the documentation of the ms/integrate file. It is not really trivial to implement a FAST algorithm ...

Define the json format

I could make sense that it is an object because a unique name is required.

Would make easier to
addSerie(name, serie)
deleteSerie(name) {
delete this.series[name]
}
...

Get inspired by

http://www.cheminfo.org/Image/Biology/Counting_plates/index.html

And look in the init script