Re: [base] spot images error
Thanks, that was the problem. It is working now. One quirk/bug I came across is that after re-creating spot images for a bioassay, the images from the original creation still remain. Logging in and out of Base does not correct this. However, refreshing the webpage (which automatically logs you out) and then logging back in causes the new images to be displayed. Dion Agriculture and Agri-Food Canada - Agriculture et Agroalimentaire Canada -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Nicklas Nordborg Sent: Thursday, March 27, 2008 2:59 AM To: BASE ML Subject: Re: [base] spot images error Lepp, Dion wrote: I am uing BASE 2.4.5 (build #3894; schema #40) Dion Agriculture and Agri-Food Canada - Agriculture et Agroalimentaire Canada -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Nicklas Nordborg Sent: Wednesday, March 26, 2008 1:16 PM To: BASE ML Subject: Re: [base] spot images error Lepp, Dion wrote: I'm trying to create spot images from a tiff file (single layer from Scanarray software) and keep receiving an error with the following stack trace. The image is 5 um, and I'm using the parameters given in the .gpr file (xscale= 5, yscale=5, xoffset= 1155, yoffset = 1577) and a spot size = 32. I've tried a number of variations on these values with no luck. Any ideas? It happens because there is at least one spot which has a null value in it's X coordinate. It is not possible to extract the spot images unless the data contains X and Y coordinates and those spots should of course be ignored. We'll fix this in 2.6.2. /Nicklas - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketp lace ___ The BASE general discussion mailing list basedb-users@lists.sourceforge.net unsubscribe: send a mail with subject unsubscribe to [EMAIL PROTECTED] - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace ___ The BASE general discussion mailing list basedb-users@lists.sourceforge.net unsubscribe: send a mail with subject unsubscribe to [EMAIL PROTECTED]
Re: [base] spot images error
Lepp, Dion wrote: Thanks, that was the problem. It is working now. One quirk/bug I came across is that after re-creating spot images for a bioassay, the images from the original creation still remain. Logging in and out of Base does not correct this. However, refreshing the webpage (which automatically logs you out) and then logging back in causes the new images to be displayed. I think that is your browser that caches the images. The reason that logging in and out helps is that the session ID is part of the URL to the images which means that they are reloaded by the browser. /Nicklas - Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace ___ The BASE general discussion mailing list basedb-users@lists.sourceforge.net unsubscribe: send a mail with subject unsubscribe to [EMAIL PROTECTED]
[base] spot images error
I'm trying to create spot images from a tiff file (single layer from Scanarray software) and keep receiving an error with the following stack trace. The image is 5 um, and I'm using the parameters given in the .gpr file (xscale= 5, yscale=5, xoffset= 1155, yoffset = 1577) and a spot size = 32. I've tried a number of variations on these values with no luck. Any ideas? java.lang.NullPointerException at net.sf.basedb.core.SpotImages.createSpotImages(SpotImages.java:487) at net.sf.basedb.plugins.SpotImageCreator.run(SpotImageCreator.java:148) at net.sf.basedb.core.PluginExecutionRequest.invoke(PluginExecutionRequest.java:89) at net.sf.basedb.core.InternalJobQueue$JobRunner.run(InternalJobQueue.java:421) at java.lang.Thread.run(Thread.java:595) Thanks, Dion Dion Lepp, M.Sc. Guelph Molecular Biology Research Unit Agriculture and Agri-Food Canada/Agriculture et Agroalimentaire Canada Telephone/Téléphone: 519-780-8090 Facsimile/Télécopieur: 519-824-2600 93 Stone Road West Guelph, Ontario N1G 5C9 [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] image001.jpg- Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace___ The BASE general discussion mailing list basedb-users@lists.sourceforge.net unsubscribe: send a mail with subject unsubscribe to [EMAIL PROTECTED]
[base] SPOT IMAGES Problem found
Hi folks, I finally managed to identify the problem. As i said in my previous email i tried a tif image with only a signe channel: [EMAIL PROTECTED] SPOTS]$ ls * -rw-r--r-- 1 david bioinfo 9,1M Jan 23 12:19 532.tif -rw-r--r-- 1 david bioinfo 9,3M Jan 23 12:20 635.tif -rw-r--r-- 1 david bioinfo 20M Jan 23 12:21 All.tif CY5 image [EMAIL PROTECTED] SPOTS]$ identify 635.tif 635.tif TIFF 1604x5108 1604x5108+444+100 PseudoClass 65536c 16-bit 9.20849mb Cy3 image [EMAIL PROTECTED] SPOTS]$ identify 532.tif 532.tif TIFF 1604x5108 1604x5108+444+100 PseudoClass 65536c 16-bit 9.02034mb [EMAIL PROTECTED] SPOTS]$ identify All.tif All.tif[0] TIFF 550x1429 550x1429+0+0 PseudoClass 65536c 16-bit 19.8007mb 1.540u 0:09 All.tif[1] TIFF 550x1429 550x1429+0+0 PseudoClass 65536c 16-bit 19.8007mb 1.450u 0:09 All.tif[2] TIFF 1604x5108 1604x5108+444+100 PseudoClass 65536c 16-bit 19.8007mb 1.380u 0:09 All.tif[3] TIFF 1604x5108 1604x5108+444+100 PseudoClass 65536c 16-bit 19.8007mb 0.700u 0:08 The experiment is a CY5 channel scan. Well the size of the single CY5 tif image is (1604*5108). The same for the CY3 image. The multi-image tif file returns 2 sizes (1604*5108 and 550*1429). With that being said i would say the problem comes from base to handle multi-file generated by genepix. It seems that base is reading the tif file sequentially and therefore get wrong coordinates for the Cy5 channel (base takes 550*1429 whereas the correct coordinates are 1604*5108). I guess this is the problem. I would say that at some point base needs to make the difference from genepix multiple-tiff files and single tif files. When I use the single CY5 image everyting works fine !!! What do you think ? - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV___ The BASE general discussion mailing list basedb-users@lists.sourceforge.net unsubscribe: send a mail with subject unsubscribe to [EMAIL PROTECTED]
Re: [base] SPOT IMAGES Problem found
david vilanova wrote: Hi folks, I finally managed to identify the problem. As i said in my previous email i tried a tif image with only a signe channel: [EMAIL PROTECTED] SPOTS]$ ls * -rw-r--r-- 1 david bioinfo 9,1M Jan 23 12:19 532.tif -rw-r--r-- 1 david bioinfo 9,3M Jan 23 12:20 635.tif -rw-r--r-- 1 david bioinfo 20M Jan 23 12:21 All.tif CY5 image [EMAIL PROTECTED] SPOTS]$ identify 635.tif 635.tif TIFF 1604x5108 1604x5108+444+100 PseudoClass 65536c 16-bit 9.20849mb Cy3 image [EMAIL PROTECTED] SPOTS]$ identify 532.tif 532.tif TIFF 1604x5108 1604x5108+444+100 PseudoClass 65536c 16-bit 9.02034mb [EMAIL PROTECTED] SPOTS]$ identify All.tif All.tif[0] TIFF 550x1429 550x1429+0+0 PseudoClass 65536c 16-bit 19.8007mb 1.540u 0:09 All.tif[1] TIFF 550x1429 550x1429+0+0 PseudoClass 65536c 16-bit 19.8007mb 1.450u 0:09 All.tif[2] TIFF 1604x5108 1604x5108+444+100 PseudoClass 65536c 16-bit 19.8007mb 1.380u 0:09 All.tif[3] TIFF 1604x5108 1604x5108+444+100 PseudoClass 65536c 16-bit 19.8007mb 0.700u 0:08 The experiment is a CY5 channel scan. Well the size of the single CY5 tif image is (1604*5108). The same for the CY3 image. The multi-image tif file returns 2 sizes (1604*5108 and 550*1429). With that being said i would say the problem comes from base to handle multi-file generated by genepix. It seems that base is reading the tif file sequentially and therefore get wrong coordinates for the Cy5 channel (base takes 550*1429 whereas the correct coordinates are 1604*5108). I guess this is the problem. Yes, this explains everything. BASE uses the images in the order they are layered in the TIFF file. I don't know why there are two extra layers i your file. Our multi-image files only have two layers, one for each channel. Maybe the two smaller images are some kind of scaled-down preview images? It would explain that some of your spot images contained a lot of small spots and also that some are totally black since the coordinates fall outside the smaller image. I would say that at some point base needs to make the difference from genepix multiple-tiff files and single tif files. When I use the single CY5 image everyting works fine !!! What do you think ? I would recommend that everyone that has a problem with multi-layered image files also try with single-layer image files. It also seems like you save some disk space if you are using single-layer image files. /Nicklas - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ The BASE general discussion mailing list basedb-users@lists.sourceforge.net unsubscribe: send a mail with subject unsubscribe to [EMAIL PROTECTED]
[base] Spot images
Hello, Following the advise given to emanuel for the generation of the spots, I went through the example but could'nt manage to create the spots. Do i have to submit three time the tiff file to create the spots in the editor (it's asking for jpeg files°. Anyway the jpeg files didn't work, i was getting Bad endianness tag (not 0x4949 or 0x4d4d). The tiff file worked out by giving only one tiff file to the red channel. However the images generated to not show the correct spot, even after setting the scale and offsets according to the definitions in the gpr file (given by PixelSize and ImageOrigin). The images are black squares and sometimes they are some spots in the same image. Has someone created spots from a gpr file. Could this person detailed: - What type of file he used (tiff or jpeg). Jpeg didn't work for me (Bad endianness tag (not 0x4949 or 0x4d4d)) - What file it did give to red, green and blue channel (if only one channel has been used do you need to fill the others) thanks david - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV___ The BASE general discussion mailing list basedb-users@lists.sourceforge.net unsubscribe: send a mail with subject unsubscribe to [EMAIL PROTECTED]
Re: [base] Spot images
david vilanova wrote: Hello, Following the advise given to emanuel for the generation of the spots, I went through the example but could'nt manage to create the spots. Do i have to submit three time the tiff file to create the spots in the editor (it's asking for jpeg files°. Anyway the jpeg files didn't work, i was getting Bad endianness tag (not 0x4949 or 0x4d4d). To create spot images you need the TIFF files. The tiff file worked out by giving only one tiff file to the red channel. However the images generated to not show the correct spot, even after setting the scale and offsets according to the definitions in the gpr file (given by PixelSize and ImageOrigin). The images are black squares and sometimes they are some spots in the same image. This is an indication of incorrect values for one or more of the scale/offset/spot size values. It could also be that the image doesn't match the raw data set. Our test data has a GPR file with the following headers: PixelSize=5 ImageOrigin=0, 0 Thus, the X and Y scale is 5 and the X and Y offset is 0. The spot size is not among the headers, but for this data 32 works fine. Has someone created spots from a gpr file. Could this person detailed: - What type of file he used (tiff or jpeg). Jpeg didn't work for me (Bad endianness tag (not 0x4949 or 0x4d4d)) - What file it did give to red, green and blue channel (if only one channel has been used do you need to fill the others) You can choose to give the file to whichever color you like the best. If you give it to all three colors the spot images will be grey. With two-channel data you should give one of the images to one color and the other to another color. If both channels are stored in the same image file, give the same file to two colors. BASE will pick the first channel for the first color and the second channel for the second color. /Nicklas - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ The BASE general discussion mailing list basedb-users@lists.sourceforge.net unsubscribe: send a mail with subject unsubscribe to [EMAIL PROTECTED]
Re: [base] Spot images follow-up
david vilanova wrote: Hello, I have tried with: PixelSize=10 ImageOrigin=4440,1000 I have double checked the image within genepix and the coordinates (expressed in µm) are correct (gpr file matches with the tiff file). 1µm correspond to 10pixels in genepix. I think it is the other way around. Each pixel is 10µm in size. See http://www.moleculardevices.com/pages/software/gn_genepix_file_formats.html#gpr X Offset: 4400 Y Offset: 1000 X and Y scale: 10 Spot size: try 32; this is given in pixels However the spot image is not the expected one. Still getting black squares and sometimes getting spots but a large set of spots instead of only one. Is a zoom applied to the spot that should be seen , otherwise i gets hards to view !?? As stated before the pixel coordinates are calculated as: pixelX = (rawX - offsetX) / scaleX and the same for the Y coordinates. It looks like soemthing is wrong and i don't know where to look ? Could I get your gpr file and tiff image to try and see how it works ?? If you go to http://base.thep.lu.se/browser/trunk/src/test/data you will find everything you need. The file 'spots.tif' is the (two-channel) image and the 'test.rawdata.import.txt' is the corresponding raw data set. /Nicklas - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.phpp=sourceforgeCID=DEVDEV ___ The BASE general discussion mailing list basedb-users@lists.sourceforge.net unsubscribe: send a mail with subject unsubscribe to [EMAIL PROTECTED]
