Mass calculation of ketohexoses

[tiskapot]

Hi

Is there a reason why the mass of the ketohexoses (fructose, psicose, sorbose, tagatose) is different from the mass of hexoses (like glucose, galactose)? To the best of my knowledge, the masses should be the same (180.156 g/mol) ...

from glypy import monosaccharides
sugars = ['Glc', 'Gal', 'Fru', 'Psi', 'Sor', 'Tag']

for sugar in sugars:
    mono = monosaccharides[sugar]
    print(sugar)
    print('> m_mono: %f' % mono.mass(average=True))
    print('> m_avg:  %f' % mono.mass(average=False))
    print('')

Glc
> m_mono: 180.156134
> m_avg:  180.063388

Gal
> m_mono: 180.156134
> m_avg:  180.063388

Fru
> m_mono: 178.140252
> m_avg:  178.047738

Psi
> m_mono: 178.140252
> m_avg:  178.047738

Sor
> m_mono: 178.140252
> m_avg:  178.047738

Tag
> m_mono: 178.140252
> m_avg:  178.047738

Thanks in advance for the reply!

[mobiusklein]

Thank you for reporting this.

I hadn't encountered any structures with these before. You're right that they shouldn't be different. If I had to guess, its because I didn't have a reference for the influence of the keto modification.

I'll look into fixing it within the next 24 hours.

[mobiusklein]

After reading a bit, it looks like ketoses like Fructose simply move the ketone group from C1 to another carbon on the backbone, and the carbonyl that normally appears at C1 is replaced implicitly with a -OH.

I was treateing the keto modification as if it meant the formation of any ketone, and not replacing the carbonyl.

[mobiusklein]

I've fixed support for ketoses, and made sure they work as expected with iupac and wurcs modules too. Changes should be in commit d6a46eb518d8fe82b3194059f058df8f98af1893. Please let me know if something still doesn't look right.

I haven't gotten around to adding updating the SNFG rendering for these yet. Is that something you think you'll need soon?

[tiskapot]

Looks good to me. Thanks a lot for the quick fix!

Updating the SNFG rendering would be nice, but for the moment I can do without.

Regarding the SNFG, I personally like the Oxford notation where the angle between monosaccharides designates the linkage.

So far the only tool I came across that implements it, is SugarSketcher, but input is only through the GUI and it doesn't allow a free end. So, integrating the Oxford notation in glypy would be a nice enhancement :)

[mobiusklein]

I have implemented a crude version of Oxford notation's layout algorithm, but not the symbol nomenclature itself, already. Just pass layout='topological' to plot. It's not always perfect about arranging distant branches so they do not overlap, but for small to medium N-glycans at least I know it works well enough.

[tiskapot]

Thanks! I'm just starting to use the glypy package and I hadn't figured out the arguments of the plot function yet. I ran a quick test for the (Sialyl-)Lewis X and A antigens, but the the 2 and 4 linkages don't seem to have the right angle.

Oxford layout

glypy output

[mobiusklein]

Looks like the special case logic that controls how Fucose is laid out under CFG/SNFG might still be influencing the topological layout.

I'll create a new issue to cover the layout problem. Is it okay to close this issue on the ketose mass calculation?

[tiskapot]

yep and thanks again