Wikipedia:Articles for deletion/Genome-wide complex trait analysis
Appearance
[Hide this box] New to Articles for deletion (AfD)? Read these primers!
- Genome-wide complex trait analysis (edit | talk | history | protect | delete | links | watch | logs | views) – (View log · Stats)
- (Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL)
This is page is an astounding work of original research and synthesis, assembling a bunch of primary sources and unsourced content into a review article. To make this a Wikipedia article it would have be completely done over. It reflects a ton of work and kudos for that, but it doesn't belong here per WP:OR and WP:NOT. Maybe Wikiversity.
I am also nominating the following page, for the same reason: Genetic correlation (edit | talk | history | protect | delete | links | watch | logs | views)
-- Jytdog (talk) 10:23, 2 December 2018 (UTC)
- Note: This discussion has been included in the list of Medicine-related deletion discussions. Jytdog (talk) 10:25, 2 December 2018 (UTC)
- Keep. I am the primary author of these two articles. Jytdog is engaged in a hostile nomination due to my comments on his other genetics AfD, Wikipedia:Articles for deletion/Designer baby (2nd nomination). (We have never interacted before that I can recall, but he has a history of this sort of thing.) His comments here are no more correct than his claims on that AfD; if you look at the 2 articles, you will see scores of citations of reviews, systematic reviews, textbooks, and meta-analyses and that they are scrupulously cited, to a fault. --Gwern (contribs) 21:12 2 December 2018 (GMT)
- I am sorry you feel that it is hostile. These two pages are what they are; there are loads of primary sources per WP:MEDDEF, and these pages don't belong in WP per the nomination. You should publish these pages somewhere else as your own reviews of the field. We will see what others say. Jytdog (talk) 21:46, 2 December 2018 (UTC)
- Optimistic keep-but-crop-heavily This is indeed an impressive piece of work, but I agree that it is too heavy on the synthesis of primary material to sit well with the sourcing balance that is expected for a WP article. It's a shame because there's easily enough review and meta-analysis material in there to pass muster for a topic - e.g. [1], [2], [3], [4], [5], [6]. BUT the synthesis, well-referenced as it may be, is very heavy here.
- Starting from the end, most of the primary source overkill is located in the "Traits" section - that, I believe, is really out of scope, what with pulling together almost 200 (!) primary research results. Personally I quite like having this available - looks like a great resource; but the arguments against it do have grounding in policy. - Sections "Benefits", "Disadvantages" and "Interpretation" seem well reasoned, but they are reasoned - they consist of conclusions drawn by the author from an assembly of primary material. Again, no good for an encyclopedia. Section "Implementations" I don't find troublesome because it just lists available software approaches, and boy do we ever have a tradition of that. "History" I think works because it's mostly a chronological assembly of material, which doesn't really fall under "synthesis". The lede reads well but I assume that much of it would have to be removed if one took away the original reasoning sections.
- Overall, yes, most of the meat of this article is too synthetic for local consumption. I wish it could be cut down to what's reasonable, though, instead of deleted, because there's a lot of effort and expertise here that would be a shame to just chuck out. If the author could just get this entire thing published as a review article in a journal, then one could go to town with that source... --Elmidae (talk · contribs) 14:43, 3 December 2018 (UTC)
- Add: I've only just looked at Genetic correlation, and it's the same in green. Great work, wrong venue, cutting it down would be a lot of effort but may be worth it.
References
- ^ Eric Turkheimer ("Still Missing", Turkheimer 2011)
- ^ "Top 10 Replicated Findings From Behavioral Genetics", Plomin et al 2016
- ^ "Meta-analysis of the heritability of human traits based on fifty years of twin studies", Polderman et al 2015
- ^ "Still Chasing Ghosts: A New Genetic Methodology Will Not Find the 'Missing Heritability'", Charney 2013
- ^ "Knowns and unknowns for psychophysiological endophenotypes: Integration and response to commentaries", Iacono et al 2014
- ^ Krishna Kumar, Siddharth; Feldman, Marcus W.; Rehkopf, David H.; Tuljapurkar, Shripad (2016-01-05). "Limitations of GCTA as a solution to the missing heritability problem". Proceedings of the National Academy of Sciences of the United States of America. 113 (1): E61–70. doi:10.1073/pnas.1520109113. ISSN 1091-6490. PMC 4711841. PMID 26699465.
- --Elmidae (talk · contribs) 14:47, 3 December 2018 (UTC)
- User:Elmidae would you be open to draftification rather than deletion? That would be OK with me; this should not be in mainspace until it is cleaned/done over... Gwern what do you think? Jytdog (talk) 15:58, 3 December 2018 (UTC)
- That would seem reasonable to me, but would of course depend on Gwern's interest in reworking the articles. --Elmidae (talk · contribs) 16:40, 3 December 2018 (UTC)
- Obviously, I disagree. There would be no point to attempting to get this published as it is merely compiling what has already been published; everything is drawn from the provided references, whether that's Neale's textbook, Plomin's textbook, Lee's review, the many phenome papers, the Cheverud's conjecture papers, and so on and so forth. It's all heavily cited, to a fault. Everyone here is freely claiming that this is 'all synthesis' or OR, and yet, no one is giving examples of what they think the OR is.
- As far as Vrie0006's comment goes: we already discussed it, and I support moving the trait references out to the relevant trait articles as part of their genetic coverage. The information about the genetic heritability or correlations about most traits on WP is in terrible shape, and that's part of why I was compiling the section, as a draft for adding to other articles. I do not have time now to do so myself (and even less time if I must deal with spurious AfDs launched to retaliate against me for my comments on another AfD); Vrie was, however, less interested in helping with this and more interested in deleting them wholesale, and of course I did not support that idea. --Gwern (contribs) 16:50 4 December 2018 (GMT)
- Gwern, as a concrete example - the start of the "Interpretation" section:
GCTA estimates are often misinterpreted as "the total genetic contribution", and since they are often much less than the twin study estimates, the twin studies are presumed to be biased and the genetic contribution to a particular trait is minor.[35] This is incorrect, as GCTA estimates are lower bounds.
A more correct interpretation would be that: GCTA estimates are the expected amount of variance that could be predicted by an indefinitely large GWAS using a simple additive linear model (without any interactions or higher-order effects) in a particular population at a particular time given the limited selection of SNPs and a trait measured with a particular amount of precision. Hence, there are many ways to exceed GCTA estimates
- followed by examples based on half a dozen primary studies, finished off with an apparently novel example calculation. That is your synthesis, and your original composition of previously unrelated material into an argument structure. Which is fine for a journal article, but not for a Wikipedia article. --Elmidae (talk · contribs) 17:18, 4 December 2018 (UTC)
- User:Elmidae would you be open to draftification rather than deletion? That would be OK with me; this should not be in mainspace until it is cleaned/done over... Gwern what do you think? Jytdog (talk) 15:58, 3 December 2018 (UTC)
- Keep - Notable topic and more than enough references. ♟♙ (talk) 21:49, 3 December 2018 (UTC)
- Thanks for entirely missing the point, mate. Truly a "drive-by" comment. --Elmidae (talk · contribs) 08:07, 4 December 2018 (UTC)
- Draftify. Per Elmidae and Jytdog's discussion. GREML (at intro of article) does get used in genetic analyses like genome wide association mapping to the point I even recognize it from skimming papers, but the state of the article just isn't ready for mainspace. The topic can have very concise content and be notable, but I also can't vote keep due to all the synthesis, etc. mentioned above. Kingofaces43 (talk) 03:53, 4 December 2018 (UTC)
- Note: This discussion has been included in the list of Science-related deletion discussions. Kingofaces43 (talk) 03:56, 4 December 2018 (UTC)
- Draftify. I'm not sure what that means but FOR SURE this page should not be merged with Genome-wide Complex Trait Analysis for many reasons. First and foremost, genetic correlations should not be equated in any way with GCTA. GCTA is a software tool, not a scientific concept. Genetic correlations can be calculated in many, many ways including pedigrees, twins, adoptees, GCTA, and others still. However, and as I've said on the talk page, the article as written leaves much to be desired. For example, the lists of trait pairs, as much work as it represents, should be entirely deleted IMO, with perhaps a few notable examples. The original research on this page is not a resource for the field, and such a resource is not effective as a wikipedia page. I offered to help in the past but found the page's primary author was highly protective, which is understandable. Vrie0006 (talk) 14:53, 4 December 2018 (UTC) P.S.: Sorry, missed the point here that there was no discussion of merge, but rather that both pages were nominated for deletion. I would still say draftify, but draftify here means deleting tons of content. I'm OK with that. Vrie0006 (talk) 14:54, 4 December 2018 (UTC)
- Draftify both per Elmidae's comments above. These look to be the start of a viable review article, but that's not what we publish. However, the problem may not be intrinsic to the article; that is, it might be remediable through ordinary editing. That editing should take place in the Draft area. In addition, many of the citations appear to be to preprints, which in this case violates WP:MEDRS. Links to preprints (arXiv, biorXiv, faculty websites, etc.) should be provided as supplements to journal versions. XOR'easter (talk) 15:25, 4 December 2018 (UTC)
- Could you quote the part of MEDRS which bans preprints like Arxiv or Biorxiv, which are increasingly the main publishing venues of these fields? For example, He was urged at the CRISPR conference to upload his article on the CRISPR babies to Biorxiv; which part of MEDRS would ban citing this upload? --Gwern (contribs) 16:50 4 December 2018 (GMT)
- Quoting WP:MEDRS, with emphasis added in places:
Ideal sources for biomedical information include: review articles (especially systematic reviews) published in reputable medical journals; academic and professional books written by experts in the relevant fields and from respected publishers; and guidelines or position statements from national or international expert bodies. Primary sources should generally not be used for medical content – as such sources often include unreliable or preliminary information, for example early in vitro results which don't hold in later clinical trials.
Preprints on arXiv and biorXiv are primary sources that have not undergone peer review. In some circumstances, preprints can be cited as sources for expert opinion (the relevant bit of policy being WP:SPS). In other cases, a preprint might effectively be "reviewed" in secondary sources before it is formally published. I don't think that those exceptions apply here. XOR'easter (talk) 17:16, 4 December 2018 (UTC)
- Quoting WP:MEDRS, with emphasis added in places:
- Could you quote the part of MEDRS which bans preprints like Arxiv or Biorxiv, which are increasingly the main publishing venues of these fields? For example, He was urged at the CRISPR conference to upload his article on the CRISPR babies to Biorxiv; which part of MEDRS would ban citing this upload? --Gwern (contribs) 16:50 4 December 2018 (GMT)
- Keep and improve with greater stress on the methodological problems of GCTA. --Davidcpearce (talk) 16:31, 4 December 2018 (UTC)
- Keep I've removed the absurdly long tables of correlations; if there are secondary sources talking about these they would still need to be re-written from scratch to be in a format useful to readers. I don't see any rationale for deleting the rest of the articles. power~enwiki (π, ν) 17:31, 4 December 2018 (UTC)
- Keep Per the above; also a bad-faith nomination by a user who has since left Wikipedia "under a cloud". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:30, 4 December 2018 (UTC)
- I was wondering when the Jytdog grave-dancers would turn up. This rationale is complete bullshit and lowers my opinion of you a lot, Andy. --Elmidae (talk · contribs) 18:40, 4 December 2018 (UTC)
- Keep and improve as needed. This is typical deletionist overreach (typical of this user, who has by my count nominated 162 articles for deletion!). If something is wrong with an article, fix it instead of culling it. Deletion should only be used for topics that don't meet notability requirements. Genetic correlation and genome-wide complex trait analysis are obviously important scientific concepts that merit good coverage in Wikipedia. Deleet (talk) 18:43, 4 December 2018 (UTC)
- Hey look, another one. You are entirely wrong, by the way; lack of notability is one of 14 defined deletion criteria. --Elmidae (talk · contribs) 19:57, 4 December 2018 (UTC)