Nmrpca fix egligibility

[comcon1]

Now, all the lipids having sn-1 and sn-2 tails are egligible. List is not hardcoded anymore.

---

📚 Documentation preview 📚: https://databank--448.org.readthedocs.build/

[pbuslaev]

Should this be stored elsewhere and only imported here? I believe that tail-names are universal and not class-specific.

[pbuslaev]

Also, I believe that this method can hardly be generalized when 4-tailed lipids become available. Should the Parser class have a dict of supported tail combinations

from collections import Counter
from ... import TAIL_VARS, REVERSE_TAIL_VARS  # dictionary of tail alt-names
SUPPORTED_LIPIDS = {
    2: ["sn1", "sn2"],  # List of tail names for lipids with 2 tails
}

@classmethod
def verify_lipid(cls, lipid: Lipid) -> bool:
    tails = []
    for v in lipid.mapping_dict.values():
        if v["FRAGMENT"] in REVERSE_TAIL_VARS:
            tails.append(REVERSE_TAIL_VARS[v["FRAGMENT"]])
    if len(tails) not in SUPPORTED_LIPIDS:
        return False
    return Counter(tails) == Counter(SUPPORTED_LIPIDS[len(tails)])
"""

[comcon1]

I isolated the function of counting tails because it should not belong to this module. Tail naming should be documented for various types of lipids. sn-1/2 is only for glycerolipids of non-archea. It doesnt' cover the variability. Will develop the function tomorrow according to which tests I will invent.

[comcon1]

I've made. Not exactly this way, but something like

[comcon1]

We delivered quite a large PR in the mean-while, so I rebased, force-pushed and verified integration tests.

[pbuslaev]

@comcon1 As you've mentioned, this is very specific for Amber and two tails. Should we add a NOTE or TODO to mark this bit to improve in the future?

[pbuslaev]

All looks really great. There is a small comment about adding a NOTE of TODO or whichever mark, but rather than that, I am happy with this change