snoscan - copyright 1999 by T. Lowe Release 1.0 (Septebmer 4, 2020) This program and its use to find 2'-O-methylation guide snoRNAs has been described in Lowe & Eddy, "A Computational Screen for Methylation Guide snoRNAs In Yeast", Science 283: 1168-1171, 1999. Updates of this package, and the Lowe Lab snoRNA database are available at "http://lowelab.ucsc.edu/snoRNAdb/" =================================================================== See the included file GNULICENSE for terms of use, copying, and modification of this software. NOTE: This is academic software. It is subject to change, and is not fully-featured, documented, or very user-friendly. There is no manual but you are welcome to dig around in the source code if you'd like to know more than what is described in this README. For an overview of the algorithm and search strategy, see the published reference above. =================================================================== Instructions for building the multiple versions of "snoscan" on UNIX machines: Search for 2'-O-methyl guide snoRNAs snoscan aka snoscanY - original version of snoscan trained on yeast snoRNAs, used in 1999 Science paper above snoscanH - slightly updated, trained on human C/D box snoRNAs instead of yeast (which are longer) snoscanA - updated version trained on archaeal C/D box snoRNAs First, uncompress and unpack the package 1) tar -zxvf snoscan-1.0.tar.gz 2) cd snoscan-1.0 Next, you need to build the sequence-handling library: 3) cd squid-1.5.11 Now, modify "Makefile" as necessary for your machine (most machines should not require any changes to Makefile) 5) make -this should produce the library file "libsquid.a" that is required for snoscan build 6) cd .. -you should now be back in the directory "snoscan-1.0" Now, modify this "Makefile" as necessary for your machine (most machines should not require any changes to Makefile) 7) make This should build the binaries "snoscan (snoscanY), snoscanH, snoscanA". 8) Make sure you have PERL 5.0 or greater installed on your machine 9) Check the first line of the included perl script "sort-snos" and make sure "/usr/local/bin/" is the directory where the perl program resides. If not, change "/usr/local/bin" to the directory where the perl binary is installed. You should now be ready to run snoscan. You can either run the snoscan program directly, or you can use the included shell script "scan-yeast" that: a) runs the program, using yeast rRNA methyl sites and yeast rRNA sequence with your own query sequence b) sorts the results with the perl script "sort-snos" for your convenience To run "scan-yeast" type scan-yeast NOTE: _must_ be in FASTA format I have provided three sample snoRNA sequences in the package to test the program: snR41.fa, snR60.fa, and snR65.fa. However, you can scan any FASTA file (with one or multiple sequences in them). You can even scan the entire yeast genome, but please note that this type of search takes about 6 hours and can consume up to 100 MB of disk space for the raw output (depending on the search sensitivity). Also note that several yeast snoRNAs require special parameters to either be detected (snR70, snR71) or call the right 5' and 3' bounds (U24, snR63). To scan snR41.fa for the methyl guide snoRNA, type scan-yeast snR41.fa This will produce serveral output files: snR41.fa.raw The raw snoRNA search output, unsorted snR41.fa.sortRH Search output, sorted by the hit's occurence in the query sequence, with some redundant hits removed snR41.fa.sort-all Search output, all hits sorted from best hit to worst, only snoRNA hits over 20 bits kept snR41.fa.sort-bysite Best 2 snoRNA hits predicted to guide methylation at each known 2'-O-methyl site, sorted in order of methylation site If you're feeling a little more bold, you can try the same on the fasta file "Sc-all-snos.fa" which contains all 41 S. cerevisiae methylation guide snoRNAs (snR70 and snR71 will not be detected using the default params; for snR63, use the "-d 250" parameter to call the correct 3' end). The snoRNA sequences in Sc-all-snos.fa are padded by 10 nucleotides on both the 5' and 3' ends, so hits should start at nucleotide 11. There will be a lot of suboptimal hits in the raw output file, so look in the ".sort-bysite" file to see top hits for each of the "known" rRNA methylation sites. You can also run "snoscan" directly, if you wish. See the shell script "scan-yeast" or type snoscan without any parameters to see example command-line options. If you create your own methylation sites file, be sure to follow the "Sc-meth.sites" file as a guide or template. The order and names of sequences in the rRNA file (first word following the ">" symbol) must be the same as in the meth sites file. Please note that this is a preliminary "research" version of the program, so certain features will be streamlined and improved before general, non-beta release. Also, I plan on making a web search page available by early summer 1999. If you're not a UNIX-type and can't get help setting up this program, the web search page should be the next best option. Comments and suggestions are welcome (please know, I'm writing up my thesis to defend in May '99 -- I may be slow making improvements and/or getting back to you). Todd Lowe First release: 02/19/99 Last revised: 09/04/2020