search

cheminfo / nmrium

React component to display and process nuclear magnetic resonance (NMR) spectra.
https://docs.nmrium.org
MIT License
55 stars 25 forks source link

Range picking of ethylbenzene is wrong #1027

Open lpatiny opened 3 years ago

lpatiny commented 3 years ago

image

jobo322 commented 3 years ago

At the moment we are not able to changes the default parameters of automatic-peak picking in nmrium, it is due probably to a big number of points. We need to think about a dynamic window size for GSD.

lpatiny commented 3 years ago

@jobo322 Could you test locally by hardcoding some parameters to check if it helps ?

jobo322 commented 3 years ago

we need be able to modify the some parameter for the ranges detection tool, otherwise it will not works for all situation. e.g the signal of citrate has a j-coupling of aprox 16 Hz. but the next region has several signals that we should not join as a signal. So how we should handle this peaks grouping.

image

jobo322 commented 3 years ago

I did increment the percent of intensity with respect to the top that a peak should reach to take it into the account. The new default value is 0.01 (10 %)

image

lpatiny commented 3 years ago

Not sure what means this factor. Is it based on the top of a multiplet or the global ratio in the full spectrum.

What is sure is that an heptuplet is not exceptional. A isopropyl for example. The ratio is then 1 6 15 20 15 6 1. Even a isobutyl would give the a nonaplet. So 10% does not seems like a good default value.

Concerning the following molecule the system finds 2 doublets instead of a doublet of doublet as well.

We need to add options for the automatic peak picking.

With ethyl vinyl ether it is quite wrong

image