Seqanswers Leaderboard Ad

Collapse

Announcement

Collapse
No announcement yet.
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Not sure why, but I had to add all the boost library linking: -lboost_regex -lboost_system -lboost_program_options -lboost_filesystem -lboost_iostreams to the g++ command generated by scons.

    Also, does this show individual read pileup?

    The speed is nice.

    Comment


    • #17
      Time zones are tricky! I'll try to respond to all of these:

      rskr>
      Originally posted by rskr View Post
      Scons requires python, to add to the dependencies. What is wrong with a Makefile?
      I agree with westerman: straight up Makefiles on their own don't provide any cross-platform configuration flexibility (does this system use cc? gcc? Are boost libraries specified boost_regex-mt or just boost_regex? Should I disable any of the optional libraries this user doesn't have (like libxmu for copy/paste or libdevil-ilut for screenshots)?) and either demands that the user knows how to read/modify a Makefile, or rely on something like autoconf to sort all that stuff out (and autoconf comes with its own can of reliability worms). I think it's realistically safe to assume everyone in the bioinformatics community has access to python at this point, and imho scons is more sane than autoconf. CMake is another nice alternative, but is harder on the end user I think (thus I've abstracted away the CMake build process the bamtools included with Samscope inside scons ).

      Comment


      • #18
        westerman>
        Originally posted by westerman View Post
        However I can not see the switch to specify the target sequences.
        There isn't one (by design). Target sequence information are extracted from the SAM/BAM header information, which for some reason isn't being successfully read for your Trinity generated (technically Bowtie->Samtools generated I assume?) BAM files. Like I said before, BAM format parsing is handled by the bamtools, so if the "bamtools header -in foo.bam" command works, so should samscope. That said, on the chance that it actually was bamtools that was broken, I've just released Samscope 1.6.6.3 which includes a newer version of Bamtools (1.02 -> 2.1.1), which may solve the problem if it was a weird compatibility issue. All of my test cases still work, so Bamtools still seems safe to me.

        Also, you showed some of the headers earlier, so another option is to put those headers into some file "foo.sam" and go
        Code:
        samtools reheader foo.sam broken.bam
        If that still isn't working, please send me a sample malfunctioning BAM file (like a subset of the 12GB one you've trying, made with "samtools view -b"), because I can't reproduce your error here no matter what BAM files I try.

        Incidentally, it may not be important, but the reason "apt-get source" failed is probably because you added just the "deb" line and not the "deb-src" line to your /etc/apt/sources.list, as in:
        Code:
        deb http://ppa.launchpad.net/cryptickrisp/samscope/ubuntu oneiric main 
        deb-src http://ppa.launchpad.net/cryptickrisp/samscope/ubuntu oneiric main
        But rather than manually adding deb lines, the "preferred" way for Ubuntu is to just use
        Code:
        apt-add-repository ppa:cryptickrisp/samscope
        which will add both the binary and source repositories.

        Comment


        • #19
          mchaisso>
          Originally posted by mchaisso View Post
          Not sure why, but I had to add all the boost library linking: -lboost_regex -lboost_system -lboost_program_options -lboost_filesystem -lboost_iostreams to the g++ command generated by scons.

          Also, does this show individual read pileup?

          The speed is nice.
          I'm glad you got it working! Yes it does show individual read pileup if you have a sorted BAM and index (.bai) file and you zoom in close enough (see also --readdetailmip).

          Thanks for the note about scons failing to add those boost library flags! I accidentally used some "scons shorthand" which seems to have been implemented in all of my release targets, but not in your version. For you it was probably finding the correct boost libraries but then neglecting to link with them. Version 1.6.6.4 should fix this problem.

          Comment


          • #20
            Got back to this. Finally got the source and complied bamtools. As a nod to 'rskr' I will complain "why do I have to install and use cmake ... isn't make good enough?" Anyway the results are null.

            Code:
            >bamtools header -in 001370_S1.coordSorted.bam 
            
            >
            In other words no output. Will send you an partial bam file in a couple of hours.

            Comment


            • #21
              As much as I dislike autotools, it is still better than cmake and scons. The problem with cmake/scons is that you have to install something, which is not so obvious to a newbie without the root permission. I would feel much happier if cmake/scons generate, say, configure.pl/build.py such that I can just launch the script to generate a Makefile or to directly compile, like what "./configure" does.

              EDIT: all the machines I work with do not have cmake installed, let alone scons which is less popular.
              Last edited by lh3; 04-06-2012, 09:33 AM.

              Comment


              • #22
                Thanks for writing and making samscope available. I struggled for a little bit with building it myself, and ran into mchaisso's problem (maybe?) where it seemed that the boost libraries weren't being found automatically, but then installed via the ppa route, and it "just worked!"

                The speed is pretty impressive, which should enable a lot of mouse-free operation. But there are a few operations that I'm not getting, or maybe aren't working for me. I started it with two BAM files and see them both after hitting 'l'. But then, how do I remove one or the other layer? 'L' only removes the second BAM file - so how do I get the second one without the first one? (Pressing "/" doesn't seem to do anything, or change which layer "L" removes)

                Then, how would I navigate to a particular range? Should I copy text from somewhere, in the format "chrX:100-200", to paste in? I tried this, and with that example, ended up looking at positions 1-100. Hitting C-b (ctrl-b) doesn't seem to do anything, but when I start samscope, if I hit C-v right away, I go to positions 1-46344 (not sure if I ever copied that range for this reference).

                Next, hitting "P" starts a slow cycling of the color palette, but that sometimes only changes the background of the BAM filenames, not the coverage graph. How would I permanently set a green/purple palette (for two files)? (I've heard this is one of the best choices for the range and frequency of different color-blindnesses out there) "K" and "k" don't seem to do anything for me. Also, the colors don't seem to always agree with the color legend. I'm looking at base mismatches in reads that have a khaki color right now, whereas mismatch is light green in the legend.

                When I zoom in far enough, I get a read detail view. I love the read mouse-over data, BTW, with quality profile. However I only get reads for the top layer (2nd BAM file), which I can't change. I have to remove that layer to see reads for the 1st BAM file.

                Next, I'm not sure what's going on with the Consensus colors. I didn't provide a reference file (actually, is this possible?). But in any case (or that case), it looks like bases in the Consensus row only get highlighting if there are variants in any of the reads mapped to that position. However, sometimes the highlighting is the canonical color for that base, and sometimes it isn't.

                I couldn't figure out the zoom tool for a while, but just did. Click right mouse button and drag left/right to zoom smoothly. Wow! Also, hold shift while clicking right mouse button and drag left/right and/or up/down, for horizontal and vertical zooming! Double wow! Very sleek.

                Hmm. Just had to wait a bit for reads from 1st layer to load, after removing 2nd layer with read view on. I assumed it wasn't working, and may have missed it if I hadn't waited long enough (10-15 seconds?). May be good to have a "reads loading" indicator.

                Sorry for the laundry list. This has a lot of potential. Thanks again.

                Comment


                • #23
                  re: jnfass

                  jnfass> Thanks for giving it a try and getting used to the key-based controls! I'm glad you're finding it useful, and don't mind a laundry list

                  Originally posted by jnfass View Post
                  Thanks for writing and making samscope available. I struggled for a little bit with building it myself, and ran into mchaisso's problem (maybe?) where it seemed that the boost libraries weren't being found automatically, but then installed via the ppa route, and it "just worked!"
                  Is this on a source code version >= 1.6.6.4? I haven't been able to reproduce this problem with source later than 1.6.6.4, so if you could tell me your distro version and scons version, I'd like to try to reproduce this bug.

                  Originally posted by jnfass View Post
                  The speed is pretty impressive, which should enable a lot of mouse-free operation. But there are a few operations that I'm not getting, or maybe aren't working for me. I started it with two BAM files and see them both after hitting 'l'. But then, how do I remove one or the other layer? 'L' only removes the second BAM file - so how do I get the second one without the first one? (Pressing "/" doesn't seem to do anything, or change which layer "L" removes)
                  This was a design decision keep things from getting too complicated. I figure it shouldn't be too hard to just remove the top layers with 'L' then re-add things in the order you like. I'm not sure where you got the idea for /, but I suppose making a "selectable active layer" which could be cycled with "/" would be reasonable. I'll add it too the list of features to experiment with!

                  Originally posted by jnfass View Post
                  Then, how would I navigate to a particular range? Should I copy text from somewhere, in the format "chrX:100-200", to paste in? I tried this, and with that example, ended up looking at positions 1-100. Hitting C-b (ctrl-b) doesn't seem to do anything, but when I start samscope, if I hit C-v right away, I go to positions 1-46344 (not sure if I ever copied that range for this reference).
                  That is the idea for how to jump to arbitrary positions. When you paste something in, it should show on the console what text it saw on the clipboard and how it parsed it. That may provide some enlightenment, however at the moment there seems to be a bug (at least in 1.6.6.6) involving the bookmark function where after you've made a bookmark the only thing that can be pasted is that bookmark. This will be fixed in the next version, I promise

                  Incidentally the ctrl-b bookmark function copies the current location to the system clipboard, so you can paste it into some other program to make a list of interesting locations or whatever you like.

                  Originally posted by jnfass View Post
                  Next, hitting "P" starts a slow cycling of the color palette, but that sometimes only changes the background of the BAM filenames, not the coverage graph. How would I permanently set a green/purple palette (for two files)? (I've heard this is one of the best choices for the range and frequency of different color-blindnesses out there) "K" and "k" don't seem to do anything for me. Also, the colors don't seem to always agree with the color legend. I'm looking at base mismatches in reads that have a khaki color right now, whereas mismatch is light green in the legend.
                  I was feeling pretty good that I actually avoided the classic red/green trap by making the first two colors red/blue, but I suppose green/purple is pretty color-blind stable.

                  Palette selection is unfortunately tricky, and something I'm open to suggestions for how to address. It may be necessary to really create a traditional break-out GUI with button etc to provide decent palette selection, but maybe some user configurable "palette/theme files" would be another decent approach.

                  As for the "rainbow palette" mode not cycling colors, this is a very non-obvious technical issue, but the short answer is: if you hit 'K' once to disable display lists (which make rendering faster in many situations), you'll see the palettes change constantly. You can hit 'n' to check the current display list sate.

                  Based on this comment, I think it's bad to expect users to guess what's going on with display lists. I'll come up with some compromise of performance and palette friendliness for the next version.

                  Originally posted by jnfass View Post
                  When I zoom in far enough, I get a read detail view. I love the read mouse-over data, BTW, with quality profile. However I only get reads for the top layer (2nd BAM file), which I can't change. I have to remove that layer to see reads for the 1st BAM file.
                  This is essentially correct. You may want to open a second window with the 'w' key to view both sets of reads at once.

                  Originally posted by jnfass View Post
                  Next, I'm not sure what's going on with the Consensus colors. I didn't provide a reference file (actually, is this possible?). But in any case (or that case), it looks like bases in the Consensus row only get highlighting if there are variants in any of the reads mapped to that position. However, sometimes the highlighting is the canonical color for that base, and sometimes it isn't.
                  Yes it is possible to infer the original reference sequence using the MD tags in BAM files. Thus you won't see any reference sequence for regions that have zero coverage, but areas with coverage are shown. The "highlighting" below the axis down there is admittedly confusing, so I've added an FAQ section explaining it. Essentially the top set of bars (behind the consensus text) are only drawn when there is some mutation called. These are drawn using the same mutation coloring scheme that (by default) the read detail display uses: matches are red, mismatch green, deletions blue, insertions purple, etc. (hit 'g' for the legend). The bottom row of bar graphs shows the called base frequency using the base coloring scheme (again, see the legend with 'g'). Also, I suggest you try mousing over those bases to make more sense of them

                  Originally posted by jnfass View Post
                  I couldn't figure out the zoom tool for a while, but just did. Click right mouse button and drag left/right to zoom smoothly. Wow! Also, hold shift while clicking right mouse button and drag left/right and/or up/down, for horizontal and vertical zooming! Double wow! Very sleek.
                  Exactly! I'm glad you figured that out! I'm thinking a youtube video or something explaining the interface controls would be helpful. Another thing on the todo list.

                  Originally posted by jnfass View Post
                  Hmm. Just had to wait a bit for reads from 1st layer to load, after removing 2nd layer with read view on. I assumed it wasn't working, and may have missed it if I hadn't waited long enough (10-15 seconds?). May be good to have a "reads loading" indicator.
                  Good idea! That won't be hard to add and should make it into the next version.

                  Originally posted by jnfass View Post
                  Sorry for the laundry list. This has a lot of potential. Thanks again.
                  Not at all! Thanks for all the helpful feedback!

                  Incidentally you might want to subscribe to the samscope-announce list to get announcements about new versions which will address the points in your post

                  Comment


                  • #24
                    Got it to compile and run on my Fedora 16 workstation, after installing some dependencies.

                    I like the browsing speed, but it'd be nice with a better presentation of the annotation; in that regard I find IGV way better - I'm working with RNA-Seq data and find it very useful to see individual "sub-tracks" for each transcript with graphically depicted splices...

                    It seems to have problems with splice-mapped reads (I have data mapped with GSNAP), causing a lot of debug output on the console. I can give you a dump if you're interested in debugging it...

                    Comment


                    • #25
                      samscope error

                      Hi,

                      I'm currently testing Samscope v.1.6.6.8 on the last Ubuntu distribution. But for the moment I didn't manage to go very far because I'm stuck with the following error message, which, I must confess, doesn't really speak to me:

                      Code:
                      > samscope SVEC.sorted.bam
                      SAM/BAM Input: SVEC.sorted.bam
                       Filetype 'sambam' specified for SVEC.sorted.bam
                      Search based on SVEC.sorted.bam
                      Searching in .
                      0 layer files found matching SVEC.sorted.bam
                      SVEC.sorted.bam bip sources missing. Building.
                      Opening SVEC.sorted.bam
                      Target 0: name='NC_004065.1' len=230278
                      Total length: 230278
                      Converting SVEC.sorted.bam to simple bip files
                      error: failed mapping diskbacked BIP block of size 460675: Invalid argument
                      If anyone here could help me figuring out which is the source of the problem, I would be very grateful. :-)

                      By the way, thinking that it was maybe due to a file format problem, I've also tried to download the sample files provided here : samscope.dna.bio.keio.ac.jp/wiki/index.php/Examples. But all I obtained is "404 Not Found" errors, like this one: "The requested URL /examples/3_1.adapterCut-195_genomes.bam was not found on this server." :/ Is there another way to gain access to them?

                      Thanks for your help

                      Comment


                      • #26
                        Any way to make Samscope work on windows 7 64 bit running Cygwin, at this time? Boost and GLUT and everything else is way beyond me.. I'm a MD, not a programmer..

                        Can anyone provide a small manual, with clear instructions?

                        Comment


                        • #27
                          Is SAMscope error free . I tried loading my BAM files on IGV and IGV gave me a lot of modification compared to what the view would be using Samtools TVIEW. In other words, I hope it does not change the BAM files and gives same visuals as TVIEW.

                          Comment

                          Latest Articles

                          Collapse

                          • seqadmin
                            Strategies for Sequencing Challenging Samples
                            by seqadmin


                            Despite advancements in sequencing platforms and related sample preparation technologies, certain sample types continue to present significant challenges that can compromise sequencing results. Pedro Echave, Senior Manager of the Global Business Segment at Revvity, explained that the success of a sequencing experiment ultimately depends on the amount and integrity of the nucleic acid template (RNA or DNA) obtained from a sample. “The better the quality of the nucleic acid isolated...
                            03-22-2024, 06:39 AM
                          • seqadmin
                            Techniques and Challenges in Conservation Genomics
                            by seqadmin



                            The field of conservation genomics centers on applying genomics technologies in support of conservation efforts and the preservation of biodiversity. This article features interviews with two researchers who showcase their innovative work and highlight the current state and future of conservation genomics.

                            Avian Conservation
                            Matthew DeSaix, a recent doctoral graduate from Kristen Ruegg’s lab at The University of Colorado, shared that most of his research...
                            03-08-2024, 10:41 AM

                          ad_right_rmr

                          Collapse

                          News

                          Collapse

                          Topics Statistics Last Post
                          Started by seqadmin, Yesterday, 06:37 PM
                          0 responses
                          10 views
                          0 likes
                          Last Post seqadmin  
                          Started by seqadmin, Yesterday, 06:07 PM
                          0 responses
                          10 views
                          0 likes
                          Last Post seqadmin  
                          Started by seqadmin, 03-22-2024, 10:03 AM
                          0 responses
                          51 views
                          0 likes
                          Last Post seqadmin  
                          Started by seqadmin, 03-21-2024, 07:32 AM
                          0 responses
                          67 views
                          0 likes
                          Last Post seqadmin  
                          Working...
                          X