Talk:Transcription preinitiation complex

Latest comment: 7 years ago by DennisPietras in topic "3Dish" images

Untitled

edit

there should also be a diagramatic representation of these factors.

I would name it Transcription preinitiation complex, because for e.g. translation PIC exists too! —Preceding unsigned comment added by 130.79.76.47 (talk) 08:47, 17 June 2009 (UTC)Reply

major revisions begun 2/26/2017

edit

Hi folks! I am intending to make major upgrades to this article. After having made major edits to Mediator in a "piecemeal" fashion (and being unhappy with that process), I intend to develop an upgraded version in one of my sandboxes before pasting into this page piece by piece. My sandbox is at User:DennisPietras/sandbox3. I don't advise you to put that on your watchlist, because I make LOTS of mistakes and repeatedly save minor differences. Keep this talk page on your watchlist, and as I complete subsections I'll post an annoucement here to invite you to have a look. I predict that this is going to take weeks, so don't hold your breath!   DennisPietras (talk) 06:08, 26 February 2017 (UTC)Reply

I have a first attempt at the lead and beginning of an Assembly section now on my sandbox. I also have the 2 images I made for this article there. I expect that at least one of you may find the colors garish  , but I had to work with what I could legally get. I believe that they are OK for those with color-blindness. The way I judge that is by opening the image in ImageJ and then making the image type 8-bit, which converts it to a gr(e/a)y scale. I can see what I want people to be able to see in that g-scale (I ain't no Inglish mayjur) image, so it is OK IMHO, even if garish. If you find something that needs to be corrected, feel free to do so. Thanks, DennisPietras (talk) 04:55, 1 March 2017 (UTC) @Slashme, Evolution and evolvability, and Boghog:Reply
The images would look a lot better with a white (or better yet transparent) background. Simply converting black to white leaves grey shadows. Doing the transfomation right is not trivial, but apparently it can be done in photoshop. Anyone have access to photoshop? Boghog (talk) 05:11, 1 March 2017 (UTC)Reply
I agree, but I don't have photoshop, nor am I going to pay for it! One problem might be that the "black" pixels near other colors aren't really black. That is, I tried to use the paint bucket in paint to make the black white, and there was a black rim around everything. sigh.DennisPietras (talk) 05:28, 1 March 2017 (UTC)Reply

@DennisPietras: I'm glad you copied me. I love drawing diagrams. There are a few things I could do:

  • Make the background white or transparent in the GIMP, and save as PNG (not hard, but will only look so-so - I'll do this this evening anyway.)
  • Trace them by hand in Inkscape (might look a bit nicer, but would take a bit longer.)
  • If someone can send me the structures involved as a molfile or similar, it would be a nice challenge to do them in Blender 3D using blmol. I could even just email Dr. Wu myself - he'll certainly have the structure files, and maybe he has nicer versions that he'd be willing to upload.

I'd say that we should separate the text from the images, partly to allow translation to be done without image editing, and partly to make the descriptions machine-readable. Because red/blue confusion is extremely unusual, only really occurring in monochromacy, we should be OK just naming the colours, but if you feel strongly, we could have the words "red" and "blue" coloured and bold. I don't like the colour changes in the word "transcription", though: it makes it seem as if the parts of the words that are differently coloured are significant, and it hinders legibility. --Slashme (talk) 08:28, 1 March 2017 (UTC)Reply

@Slashme: Well, at least you didn't think any of it was garish! I'm surprised that you objected to the colors in "transcription" but not "mediator". The colors I used in mediator came from the head, mid and tail colors used in the model. I figured that was an easy way of indicating which parts were mediator. Having done that, I figured that the multicolors of transcription factors would also indicate what the blobs symbolized. Regarding the black background: My suspicion is that Wu has an original bitmapped file from which it would be relatively easier to white out the black. When I capture a screen, I do all of my editing on 24 bitmapped files, with only the final step being adding 4TC (for The Commons) to the filename and converting it to jpg. Still, I suspect that pixels of black next to blue or red are blended to be dark red or dark blue by the conversion to a video file. So, I could e-mail you my bitmap files, but I don't think it would solve the problem.
Another thing to remember is that Wu found it perfectly acceptable to present the original images to a professional grouping, many of which were his peers, and so I think we should get on with more important things than worrying about the background color. However, if a consensus develops about the text colors, I am perfectly willing to make the changes on my bitmap files and upload replacements. Finally, in response to a suggestion on the media copyright help page, someone suggested that I e-mail the author to get an original file. My reply to that was that I made a personal decision not to take a moment of an NIH scientist's time dealing with images for wp. DennisPietras (talk) 19:37, 1 March 2017 (UTC) @Evolution and evolvability and Boghog:Reply

@DennisPietras:

  • The text colour: I disapprove of the text colours of "mediator" as well - I just used "transcription" as an example.
  • Whether to make the background white: The fact that Wu decided that a black background looks better on a slideshow doesn't mean that it's the right choice for Wikipedia. Wikipedia articles frequently make it into print, and square black backgrounds are not great for print. Also, the Wikipedia style is relatively clean and light, so we generally prefer white or transparent backgrounds. If you don't believe that this is a consensus position, feel free to ask around at Wikipedia:Graphics_Lab. Please do mail me your bitmapped files: that will certainly save time: davidrichfield gmail.com .
  • My experience with contacting scientists so far has been that they are often happy to have an opportunity to share their work with a wider audience, and if I can get molfiles from him, I might get a really nice result in Blender, and who knows, he might get some benefit from the interaction as well (e.g. blender files with some improvements that he didn't consider). --Slashme (talk) 20:43, 1 March 2017 (UTC)Reply

@Slashme: I just sent the bmp files. I don't have any objection to making the background white, I just don't have the ability to do it and don't think it is worth the time. Picking out the outlines of the transcription factors pixel by pixel taxed my patience to the utmost, and I came perilously close to running out of party peanuts while doing it! DennisPietras (talk) 23:14, 1 March 2017 (UTC)Reply

@DennisPietras: Thanks, got them. I'll do the editing this evening when I get home. Fortunately I use the GIMP, which uses fewer party peanuts per pixel than Paint :-D --Slashme (talk) 07:53, 2 March 2017 (UTC)Reply
So in the end I only did one this evening, because I got sidetracked trying to import this structure into Blender, and ended up hacking at someone else's Blender Python code. Probably best to get your input on what the first two pictures look like, anyway. --Slashme (talk) 22:47, 2 March 2017 (UTC)Reply
I am in awe of what you did! I had GIMP on my computer about a decade ago, but gave up on it in favor of paint. I hope I never need to white out black again, but if I do, I'm going to have to look at GIMP again. As for the stuff you were trying to import into blender- I have no clue about the program. But, I will offer this about my intention- I was going to stop this page at the formation of the open PIC. The moment the first nucleotide is incorporated should be the start of the eukaryotic transcription page, IMHO.   Thank you for the images! DennisPietras (talk) 03:18, 3 March 2017 (UTC)Reply
@Slashme:I have become concerned that GIMP is an acronym warning people like me to Gather an Immense Mass of Peanuts before using.   DennisPietras (talk) 02:19, 4 March 2017 (UTC)Reply
@DennisPietras: Thanks for the compliment!! I'll do the others now that I have your approval. I was off work with a cold for the last few days, but now I'm mostly healthy again. --Slashme (talk) 13:47, 4 March 2017 (UTC)Reply
@DennisPietras: I've done two more, but I've been researching ways to make stuff clickable in the images, so I've presented two possibilities: one with links as wikitext over an image without text, and one using image maps. Let me know what you think. I think making an SVG with nice text, and an image map to give the links and annotations on mouse hover, is probably the way to go. Also, I see that you had drawn the transcription factors rather roughly. Do you have any strong feeling on what they should ideally look like? I can do much nicer things with Inkscape than this first draft... --Slashme (talk) 15:24, 4 March 2017 (UTC)Reply
@Slashme: Rather roughly????????  I lined up the 3 TF's with the model of mediator/pol II so that they would fit in a "hand and glove" manner! THAT'S what made me keep reaching for the peanuts! Then, for what you called #3, I positioned it to cover the entirety of the DNA so that I could let the DNA bend to accomodate the fit with mediator without actually needing to visually bend the double helix. I positioned #2 on just one side of the helix. For number 1, I thought of 2 tines of a fork (bringing up thoughts of forkhead proteins!!!!) wrapping down and around the DNA. If you want to do better, go ahead, I've got peanuts to drown my sorrows. I have no idea how they are supposed to look, since they are entirely fictional, however, in general, if they look more like the mediator model, they would look better and more consistent. I am once again in awe of what you've done! I prefer the image with the caption "Highlighting some chemical groups.", but I would change it to "Highlighting some biochemical modifications." If you want to do even more, I believe the blue blob to the left of "Z" is linker histone H1, and you could link the image to that page. Then, the caption would be ""Highlighting some components and biochemical modifications.". If you are going to label theoretical Tf's, I would use letters rather than numbers, to avoid confusion between a theoretical TF2 and TFII. Finally, you are making me feel inadequate to even try to illustrate a pioneer factor penetrating compact chromatin,  , but I've got peanuts!   Thanks, DennisPietras (talk) 20:43, 4 March 2017 (UTC)Reply
@DennisPietras: Eh, what I know about DNA folding and transcription could be written in large letters on small paper, so I just kind of traced your "artists impression" in Inkscape, not knowing whether they were meant to represent any specific structure. I also just picked "TF1" to "TF3" to show that they were distinct. None of the text, labels or links that I typed were meant to be significant, just to illustrate the ways in which you could represent the images. You can really see me as your illustrator more than a serious biochemist in this context. Maybe tomorrow or Monday I'll have another crack at those pictures, with a more coherent look between the TFs and the mediator model. In the meanwhile, I've been learning how to animate a cartoonish DNA model along a curve, so watch this space for chromatin unwrapping pictures... --Slashme (talk) 21:21, 4 March 2017 (UTC)Reply

Whac-A-Mole or rather Whac a Pioneer factor

edit

Hi again folks. I have become mired in another real-life game. In my sandbox, the last sentence I added linked to Pioneer factor. So, I spent considerable time learning about pioneer factors, then looked at the wp page and gasped. I understand that I've composed some stinkers, but the images are terrible - the condensed chromatin depiction made me lol! The textual content needs correcting and updating and lots of it. I'm very reluctant to even link to the page from my PIC sandbox page. I'd like to ask for advice. Do you folks think I should make a fairly detailed description of pioneer factors and their function in the PIC article and not link to the pioneer factor page, or should I update the pioneer factor page first? My inclination, based on masochism and the desire to have more reason to eat peanuts, is to update pioneer factor page first so as to avoid duplications later on between PIC and PF articles, but I am open to being talked out of it. Also, one of the major authors in the pioneer factor research arena is a guy name Ken Zaret, and he and I were both in Fred Sherman's lab together. We weren't close friends, I haven't seen or talked to him in over 30 years, and I don't think that me mentioning his work is a COI, but I want to make sure that nobody would object. Thanks for any comments. DennisPietras (talk) 02:53, 4 March 2017 (UTC)Reply

@DennisPietras: I actually rather like the clarity of those images. They're completely unrealistic, but they show the parts neatly and cleanly. I also like peanuts, though, so what do you think of a nicer version of those pictures done in Blender 3D, with DNA shown as a double helix, some slightly more realistic-looking histones, but still not too complex in shape, and some semi-convincing transcription factors? I could even do an animation with dramatic lighting. It'll be great, the audience will love it. --Slashme (talk) 15:29, 4 March 2017 (UTC)Reply
Whoops, forgot to answer the other questions. What you don't want is to be accused of doing a WP:CFORK, so if you publish this page and it overlaps massively with the Pioneer Factor page, and contradicts it, that could be a Bad Thing™. Knowing someone who was a major researcher in an area isn't a CoI, as long as you're not egregiously promoting his work. The main test is "can you write neutrally about the topic", and in this case, I'm sure the answer is yes. --Slashme (talk) 15:37, 4 March 2017 (UTC)Reply
@Slashme: Please take a moment to look at the graphical abstract at http://www.cell.com/abstract/S0092-8674(12)01298-6 What I had my heart set on (and notice how "rough" they drew the tf's!!! I could have done that!) was to use the condensed chromatin part of your white image and have an image of a sock (boghog would have loved it! ) representing a Sox protein pushing its toe into the chromatin and distorting it a bit. Then have an image of a microphone (indicating cMyc protein) coming in next to the sock and knocking a nucleosome off completely to generate the nucleosome free region. I'm guessintg that it isn't your style  but if you want to do an animation, that would obviously be fantastic! Re: CFORK. Thanks! More reason to tackle pioneer factor and more peanuts first! DennisPietras (talk) 21:12, 4 March 2017 (UTC)Reply
An animation of a DNA strand unwrapping from a histone complex.
@DennisPietras: Sock and microphone, eh? Nice mental image, but as you guessed, not quite my style ;-] I've been working in Blender a bit, and I already have a nice cartoony DNA strand unwrapping from a histone complex that I based on a PDB structure, with the histone morphing from its normal shape with spikes holding onto the DNA to something a bit smoother, so that the DNA can slip off. I'm quite happy to add Sox and cMyc into the mix, and ideally, I'd like to grab some structures off PDB, so that they will be realistic. As I see from the Sox article, it binds to the minor groove of DNA, which is apparently how it disloges the DNA from the histone? I'm not very clear about where I should bring it into the animation, though. cMyc is basically discussed as a gene promotor: they don't mention what it does to push away the nucleosome. Based on the short video clip that I've included, you might have some more detailed suggestions on how to script it. --Slashme (talk) 19:29, 5 March 2017 (UTC)Reply
@Slashme: AMAZING VIDEO! 2 general "problems" quickly before supper. 1) You mistakenly think that wp articles are up-to-date   and, 2) can you slow down the video? I'll write more after dinner. Thanks, DennisPietras (talk) 21:46, 5 March 2017 (UTC)Reply
Sure, no problem. I'd made it quick because less frames mean less render time, but I'll leave it running when I go to bed now, and you'll have a video with twice the resolution and twice as slow tomorrow. --Slashme (talk) 23:12, 5 March 2017 (UTC)Reply
@Slashme: Well, I'm glad that you get to sleep with all you have going on! Since you are determined to make this more realisitic, I'm going to copy some images from your video and find better example proteins since the sock and microphone are out. I'll e-mail jpg's to you. I feel a little bad about encouraging you to spend so much time on this, since you have a job to do, but since you're an experienced wikipedian, I'm sure you know about the risk of WikiHomelessness, which was explained to me early on by TimothyJosephWood at the teahouose! DennisPietras (talk) 01:58, 6 March 2017 (UTC)Reply
@DennisPietras: That sounds great! I see you essentially sent me a storyboard by email, so I will do the animation this week. --Slashme (talk) 08:48, 6 March 2017 (UTC)Reply
@Slashme: Hmm...I just learned another term - storyboard. Remember, no rush! Thanks again!DennisPietras (talk) 03:12, 7 March 2017 (UTC)Reply

"3Dish" images

edit
 

Hi slashme et al. Stimulated by slashme's work on the images, I'd like to make the orange, green and grey blobs of the transcription factors look more 3Dish, like the histones. I've got both GIMP and ImageJ on my computer, and I've tried screwing around with them, but am making no progess. Can anybody please give me some hints before I exhaust all my peanut supply? Thanks, DennisPietras (talk) 03:33, 15 March 2017 (UTC)Reply