jltsiren / bwt-merge Goto Github PK

Does bwt-merge work with BWTs from non-DNA alphabets?

Given that it takes SGA's BWT format as input this would presumably require SGA to support non-DNA as well or another BWT format?

I'd like to build a full-text index of all wikipedia revisions. This is something like 100TB uncompressed and it's not clear to me if there is a better way to do this than using an approach based on BWT merging. But if there is I'd be curious to know.

Hello, I have the following problem with compilation. SDSL is already compiled in ../sdsl-lite. Do you have any idea how to fix it? Thanks.

karel@optikarel:~/github/bwt-merge$ make SDSL_DIR=../sdsl-lite/ VERBOSE=1
/usr/bin/c++ -std=c++11 -Wall -Wextra -DNDEBUG   -DVERBOSE_STATUS_INFO -pthread -O3 -ffast-math -funroll-loops -msse4.2 -DHAVE_CXA_DEMANGLE -I/usr/local/include -c bwt.cpp
In file included from support.h:28:0,
                 from formats.h:28,
                 from bwt.h:28,
                 from bwt.cpp:27:
utils.h:37:34: fatal error: sdsl/wavelet_trees.hpp: No such file or directory
compilation terminated.
make: *** [bwt.o] Error 1

Hello,

I would like to ask you if BWT-merge can merge also BWA indexes (https://github.com/lh3/bwa). It would be a very useful function.

Thanks.