Mass data calculation and presentation

Hi,


now evaluating the tool "Elemental Analysis" I found some problems and inconsistencies. But first to the background. Some years ago I visited a Mass Spectroscopy Course given by McLafferty. Then I made a lot of student courses on this topic and yesterday I had a discussion with our campus mass spectrocopy specialist on the Marvin representation of mass data.





To support my argumentation I attach two results of the tool "Elemental Analysis".


1) The meaning of "Mass" is not clear. By calculating the value with another program it is clear here is the "Average molecular chemical mass" and for Cl it is 35.4530 g/mol. The dimension should be given here.


2) The so called "Exact mass" means obviously the high resolution mass of the ion with the lowest isotopes given in the molecule (In the examples C12 and Cl35). To give these values with 9 digits is unscientific. The normal accuracy in the literature documentations is about only 6 digits.


3) What is the "Isotope formula" ? Chlorine has Cl35 (76%) and Cl37 (24%). It seems this option is only for D and T. This concerns also the other "isotope" tools.


4) What is "Dot-disconnected formula" ? I couldn't find it in the documentation for GUI handling of Marvin.





In the next post I suggest to seperate this tool into two tools "Elemental Analysis" and "Mass data". Please await this next post before answering.





Regards, Hans-Ulrich

Hi,


attached you find my suggestions for the tool "Elemental Analysis" and the new tool "Mass Data".





The results are self explaining and this is very important for a program which is installed in the internet !!! . Numerous pupils and students will use it to learn chemistry. So all menus and results should be as self explaining as possible. Always to point and to refer to the help function or the (difficult to reach) manual is impracticable, and it is not used by the normal user.


Good software is as intuitive as possible.





(The submenu "Elemental Analysis Options" is superfluous. All options on.)





The calculation results in the attached screenshots come from: http://www2.sisweb.com/mstools/isotope.htm





Regards, Hans-Ulrich

Thank your for the valuable comments!





1) The meaning of "Mass" is not clear. By calculating the value with another program it is clear here is the "Average molecular chemical mass" and for Cl it is 35.4530 g/mol. The dimension should be given here.


Yes, "mass" is the molecular weight (which is average by definition). Yes the unit might be displayed as well, but I think, that in this case, it is quite obvious.








2) The so called "Exact mass" means obviously the high resolution mass of the ion with the lowest isotopes given in the molecule (In the examples C12 and Cl35). To give these values with 9 digits is unscientific. The normal accuracy in the literature documentations is about only 6 digits.


Yes, "exact mass" is the monoistopic mass, usually interesting in mass spectrometry calculated by the most abundant isotopes of elements. In both case of the "mass" and the "exact mass" values the number of resulting digits come from the least precise mass used in the calculation.


The mass of 1H is 1.007825032, the mass of 35Cl is 34.968852707, and carbon is the reference mass. So the 9 digits in case of hexachlorocyclohexane is not unscientific.


Please let me explain it via an example of a formyl cation.





- The corresponding elemental weights of the atoms of a formyl cation are:


C: 12.0107 (4 decimals)


H: 1.00794 (5 decimals)


O: 15.9994 (4 decimals)


Sum: 29.01804





Since the smallest precision is 4 decimals, this will be rounded to 29.0180





- In case of an exact weight (for mass spectrometry) having the most frequent isotopes





12C: 12.0000000 (any decimals, this is the reference point)


1H: 1.007825032 (9 decimals)


16O: 15.994914622 (9 decimals)





The smallest precision is 9, so the exact mass will be 29.002739654





- The molecular weight of the formyl cation signed with 14C:





14C: 14.003241988 (9 decimals)


H: 1.00794 (5 decimals)


O: 15.9994 (4 decimals)


Sum: 31.010581988





Since the smallest precision is 4 decimals, this will be rounded to 31.0106








3) What is the "Isotope formula"? Chlorine has Cl35 (76%) and Cl37 (24%). It seems this option is only for D and T. This concerns also the other "isotope" tools.


Formula lists atom types in the molecule and their amount (w/w %). Isotope formula displayes isotopes as separate atom types from elemental atoms. Deuterium and Tritium can be displayed as D and T, or as 2H and 3H optionally.








4) What is "Dot-disconnected formula" ? I couldn't find it in the documentation for GUI handling of Marvin.


Dot-disconnected formula provides separate formulas for disconnected structures. For example, the formula of sodium acetate "CC(=O)[O-].[Na+]" is "C2H3NaO2", its dot-disconnected formula is "C2H3O2.Na".


You can find a brief explanation of all functions in the user's guide. You will find there the source of the used element and isotope weights as well.








5. Regarding your suggestions, I understand, that you feel, that the many elemental analysis options seem superflous (though a bit less than in case of topology analysis :) ). Do you think, think, that it would have helped if we added explanatory tooltips to the options?


The suggested mass distribution calculation can be a useful, additional calculation, so we will implement it in a future version of Elemental Analyser.

Hi Gyuri, some more comments.





To 1) Yes the unit might be displayed as well, but I think, that in this case, it is quite obvious. No. Redundancy supports clarity and security.





To 2) One example more, Bromoform CHBr3, see screenshot.


The "Marvin Exact mass" differs after the fifth decimal. Which table of element masses does Marvin use? I couldn't find it in the manual and to get the original papers is too far away.





To 3) Isotope formula displayes isotopes as separate atom types from elemental atoms. There are a lot of separate atom types in CHBr3: C12, C13, H1, Br79, Br81. Why are these not displayed here, and in "Isotope composition", respectively ?





To 4) I will try it.





To 5) The suggested mass distribution calculation can be a useful. No, not "can". It is the only way to understand those isotope mass patterns. The display of Marvin is only understandable when I go to the web page where those patterns are given in a table and graphically shown. To have the little table would be enough. And the heigher masses with low abundance could be cut off, but the main pattern should be presented.





Remember: Marvin is in the web. And it is used by everyone.





BTW: Is there any statistic how many visitors use the MarvinSketch applet ?


In the svensk word book I am using I see in time how many users look also for translations.





Regards, Hans-Ulrich

Dear Hans-Ulrich,





2. The elements. zip and isotopes. zip files of Marvin contain the weight tables. These are the original weight resources:


Atom weights: M. E. Wieser, "Atomic weights of the elements 2005 (IUPAC Technical Report)" Pure Appl. Chem., Vol. 78, No. 11, pp. 2051-2066, 2006


Isotope weights: G.Audi and A.H.Wapstra, "The 1995 update to the atomic mass evaluation" Nuclear Physics A595 vol. 4, pp. 409-480, 1995





3. I think, that you misunderstand the purpose of the isotope formula. It differs from formula only in case of isotope labelled molecules. For example the formula of deuterated benzoic acid is: C7H6O2, its isotope formula is C7H5DO2.





We are glad, that Marvin is very popular on the web, we feel the responsibility and we try provide a tool that serves the community of chemists well. However, it is not always easy to prioritize the features waiting for implementation, since chemistry is a very function-rich science. Though, we are driven by our customers needs, it often takes some time when a desired feature appears in the tools.





Regarding the number of MarvinSketch applet users it is a huge number with many trailing zeros. ;)





Best regards,





Gyorgy

Dear Gyorgy,





sorry, where are these zip-files? I cannot find it. Neither in .chemaxon nor in ChemAxon. Which folder ?





Best regards, Hans-Ulrich

If you have Marvin Beans package, the zip files are wrapped into the MarvinBeans.jar (<Marvin Beans home>/lib/MarvinBeans.jar).


Inside the jar, zip files are located in "chemaxon" directory.


If you use Marvin Applets package, the "chemaxon" subfolder of the package contains these zips.

Hi,





OK. I found it and I could "unzip" it. We will compare it with the list we are using in our MS-lab.





A lot of the problems I had, came from the deficient descriptions of the values given. In my next post, there will be a suggestion for a window "Element and Mass Analysis".





Regards, Hans-Ulrich

Hi,





I used the above mentioned "dot-disconnected" SMILES formula CC(=O)[O-].[Na+] (NaAc.smi). And I got in MarvinSketch the two ions of sodium acetate (Green arrow in the screenshot). OK, looks fine.





But then using the "Elemental Analysis" (Red arrows in the screenshot) there is a window with two parts.


In the upper part is CH3COOH instead of CH3COO- and in the lower part there is Na+.


Having CH3COOH is of course wrong.





But there is no difference between "Formula" and "Dot-disconnected formula".


And there is also no difference between "Isotope formula" and "Dot-disconnected isotope formula", but this is because "Formula" and "Isotope formula" are also equal.





What's wrong ?





May I get another example? F.e.the SMILES formula for C8H11NO.ClH mentioned in the User's Guide.





Regards, Hans-Ulrich

Thank you for pointing out this error of "single fragment mode" operation, the ionization should remain intact on the separated fragments (components).


Marvin just handles then the two components separately, so the formula and dot disconnected formula are equal for an acetate and similarly the formula and dot disconnected formula are equal for a sodium cation. Or do I misunderstand your question here?

Hi Gyuri,


after heavy discussion in other topics of your forum I would like to continue the discussion here.


Can you give me please another example for a dot.disconnected formula,


f.e. the "C8H11NO.ClH" mentioned in the "User's Guide".





BTW: In the MarvinSketch Help tool the really fine search tool finds neither "dot" nor "disconnected".





Regards, Hans-Ulrich

If a molecule contains two or more separated components like in case of some salt representations (like sodium acetate), the dot disconnected formula is constructed from the formulas of the components separated by dots (like C2H3O2.Na).


However, you can probably find better explanations for the dot disconnected formula than my ones, because it is used by CAS and STN as well.





The search engine of Marvin help is not perfect indeed. We are working on porting the help into other formats, some will probably come with more sophisticated search engine, though it will take a while to complete.

Hi Gyuri,





now I constructed another example, see the screeshot attached (May be the components react in reality with each other, but adding some acceptor and donor groups will adjust the pKa values and the salt is stable. It is a theoretical example. To calculate the properties is always interesting).





But the result is obviously wrong for the "single fragment mode".





And the nomenclature is either wrong or not clear enough, as I have already commented in the posts above.





Before I make a suggestions for this Elemental and Mass Analysis tool, I will evaluate examples with special isotopes.





BTW: It's annoying to have the text in the "single fragment mode" below the formula. One cannot scale down the window without loosing the text. Why is Marvin using different windows ?





Regards, Hans-Ulrich

HUWagner wrote:

But the result is obviously wrong for the "single fragment mode".

In some cases valence check removes the charges from the molecule. It is a known bug, we are working on the fix. Fix will be available in Marvin 5.1.4.

Quote:

BTW: It's annoying to have the text in the "single fragment mode" below the formula. One cannot scale down the window without loosing the text. Why is Marvin using different windows ?

The MarvinView component that displays multiple molecules places the text below the formula.





Regards,


Zsolt