Talk:Structure validation

This page is now under construction, and would welcome edits. After discussion, it has been named just "Structure validation", with a redirect from "Macromolecular structure validation". Dcrjsr (talk) 17:12, 6 April 2012 (UTC)

Readability
I strongly recommend the writers read Make technical articles understandable and then write the lead and introduction for a broader audience. RockMagnetist (talk) 17:53, 6 April 2012 (UTC)

I rewrote the lead in an attempt to make it more readable. I tried to change the content as little as possible, with one exception. I was very skeptical of the claim that large numbers of parameters made validation easier. Quite the reverse, I would expect. Validation probably works because crystallography and NMR are sensitive to very different physical parameters. Is that correct? RockMagnetist (talk) 18:51, 6 April 2012 (UTC)


 * Thanks very much for the quick look and prompt action! Actually, altho joint methods are coming along, nearly all such models are derived from just one of the experimental techniques.  I should have said that millions of individual observations, rather than parameters, make it possible - because there are rules that govern their consistency with each other, as well as between them and the coordinates, and between the coordinates and our knowledge of physics and chemistry.  I'll try to reword that first paragraph a bit further, to get it correct but keep the good changes you've made to clarity, shorter sentences, etc.  Dcrjsr (talk) 20:08, 6 April 2012 (UTC)


 * You're welcome. I suspected that I didn't know the whole story, but it seemed worth having a first try at it. RockMagnetist (talk) 20:56, 6 April 2012 (UTC)

It should be noted in introduction that the entire issue of "validation" comes from the relatively low resolution/precision of macromolecular structures determined by crystallography or NMR. This is not really an issue for small molecules whose structures are determined by direct crystallographic methods. My very best wishes (talk) 22:15, 6 April 2012 (UTC)


 * I had a go at putting something of that sort in the intro, but found it worked better to modify the small-molecule paragraph that starts the history section. Dcrjsr (talk) 23:02, 6 April 2012 (UTC)


 * The point here are not small molecules, but telling something like this: Validation of experimental structures of biological macromolecules is important because they are determined with a relatively low precision (around +-0.3 A for for a good resolution crystal structure and +-1.5 A for solution NMR models ?). Let's also check if I understood this correctly: "The validation has three aspects: 1) checking on the validity of the thousands to millions of measurements in the experiment..." Are we talking about evaluating quality of data prior to generating models? For example, one can not expect a good quality all-atom structure from 6A resolution EM data, orientational constraints by solid state NMR or IR data. My very best wishes (talk) 14:08, 7 April 2012 (UTC)

More on readability: The historical summary is pretty technical. It might be better to have the history at the end of the article, or at least preceded by an Introduction section. RockMagnetist (talk) 23:15, 6 April 2012 (UTC)


 * I see the point. I've made the wording less technical in many places, and will work on that further as this goes along, since it would be nice to feel we could keep it at the beginning. Dcrjsr (talk) 00:46, 7 April 2012 (UTC)

Citation template
The explanation of my recent edit may not make sense, so I'll explain it here: The template harvnb is very handy for compacting repeated references. You put the reference down between Refbegin and Refend and insert harv. Then when you click on a citation like, it takes you down to the full reference. The syntax takes a little getting used to though. RockMagnetist (talk) 21:16, 6 April 2012 (UTC)


 * I did the best I could with the edit conflict, and I think actually nothing got lost from your reformatting (please check on that). I like the 2-column ref list and the neat way of doing the chapter 13 note. I presume that either one puts the complete reference info in the Refbegin and Refend section (if it'll need the harv functionality), or else just in its first citation (as I was doing before)?


 * It's probably a good idea to use full first names for the book authors, but for journal articles I'd really rather stick with the concise "Engh RA, Huber R, ..." style in one list, because there will be quite a lot of refs in here with many, many authors (like the current Read et al with 17! Is that OK with you? Dcrjsr (talk) 21:49, 6 April 2012 (UTC)


 * You missed a couple, but it only took a moment to fix. FYI, the Refbegin and Refend is just a formatting block so the references are in slightly smaller text. You're certainly welcome to format the references how you like – Wikipedia doesn't have a single recommended style – as long as it's consistent. I'm just showing you some options. With your more concise style, a different approach using the harvid template for anchors will be needed. I'll do one that way as an illustration. RockMagnetist (talk) 22:21, 6 April 2012 (UTC)


 * Another word of explanation: Unfortunately, the default citation anchors only work if you use last1, first1, etc., for the author names instead of authors. It's probably easier to use harvid. For example, if you have the citation, you can use |ref= in your reference. Just change harvnb to harvid. RockMagnetist (talk) 23:11, 6 April 2012 (UTC)


 * I moved Engh&Huber out of the Further reading, since it's a short & specialized paper. Hope I did it correctly! Dcrjsr (talk) 00:47, 7 April 2012 (UTC)