churchill-lab / emase-zero

C++ Implementation of EMASE

EMASE-Zero is a C++ implementation of EMASE, Expectation-Maximization (EM) algorithm for Allele Specific Expression (ASE). Unlike other approaches, we take into account the nested relationship of isoforms and alleles under genes when we disambiguate multiply aligning reads, which improves downstream analyses performed at the gene level. In combination with alntools (an alignment file processor that generates compressed three-dimensional incidence matrix), Zero estimates the expected read counts fast, over 10 times faster than RSEM. Zero also generalizes our fast hierarchical EM to any decent alignment strategies. The original python implementation is available at https://github.com/churchill-lab/emase.

• Free software: MIT license
• Full documentation: http://churchill-lab.github.io/emase-zero/

emase-zero's Issues

Struggling to get started

Hi, I am finding difficult to get started with emase-zero. I am following this protocol suggested on the webpage

$ bowtie -q -a --best --strata --sam -v 3 -m 100 bowtie.transcriptome my_sample.fastq | samtools view -bS - > my_sample.bam
$ alntools bam2ec -t transcripts.lengths.tsv -c 8 my_sample.bam my_sample.bin
$ emase-zero -m 2 -b my_sample.bin -o emase-zero.quantified -l transcript.lengths.tsv -g gene2transcripts.tsv

My doubts are:

• where does the bowtie.transcriptome file comes from? is it bowtie index for the reference transcriptome or is it for a personalised transcriptome?
• do I need to do I incorporate a vcf with known genetic variation for each individual? to make the a personalised transcriptome, i.e., bowtie.transcriptome used in the first step?

Do you have working example with some test files that I could follow?
Thanks a lot in advance.

Regards,

Inti

Enquiring about presence of maternal and paternal allele in the transcript-lengths file

In the GitHub page http://churchill-lab.github.io/emase-zero/ you mentioned using transcript-lengths file which contained information about the transcripts being maternal or paternal (e.g. ENSMUST00000000001_M (maternal) and ENSMUST00000000001_P (paternal)).
Can you please describe how did you identify the transcripts with maternal and paternal allele?
Did you use any particular method to identify them?

Support for 10x Genomics data?

Hi,
I have a client interested interested in single cell ASE based on the 10x Genomics platform. The default Cell Ranger pipeline (using STAR 2.5) is not capable of resolving ambiguously mapped reads across a diploid genome and she is in need of another option. Any suggestions?
Thanks in advance,
Chris

Issues compiling emase-zero

Hi,

I am running into a problem when I try to compile emase-zero. I get the following error:

g++ -lz -o emase-zero alignment_incidence_matrix.o sample_allelic_expression.o alignment_import.o emase.o
alignment_import.o: In function `getBinFormat(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >)':
alignment_import.cpp:(.text+0x404): undefined reference to `gzopen'
alignment_import.cpp:(.text+0x40f): undefined reference to `gzrewind'
alignment_import.cpp:(.text+0x421): undefined reference to `gzread'
alignment_import.cpp:(.text+0x429): undefined reference to `gzclose'
alignment_import.o: In function `readIntFromFile(gzFile_s*, std::basic_ifstream<char, std::char_traits<char> >&)':
alignment_import.cpp:(.text+0x8f7): undefined reference to `gzread'
alignment_import.o: In function `readCharFromFile(gzFile_s*, std::basic_ifstream<char, std::char_traits<char> >&)':
alignment_import.cpp:(.text+0x977): undefined reference to `gzread'
alignment_import.o: In function `fileSeek(gzFile_s*, std::basic_ifstream<char, std::char_traits<char> >&, long)':
alignment_import.cpp:(.text+0x9f2): undefined reference to `gzrewind'
alignment_import.o: In function `loadNFromBin(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, AlignmentIncidenceMatrix&, int)':
alignment_import.cpp:(.text+0xcfe): undefined reference to `gzopen'
alignment_import.cpp:(.text+0x1006): undefined reference to `gzrewind'
alignment_import.cpp:(.text+0x1016): undefined reference to `gzseek'
alignment_import.cpp:(.text+0x103b): undefined reference to `gzread'
alignment_import.cpp:(.text+0x1064): undefined reference to `gzread'
alignment_import.cpp:(.text+0x1087): undefined reference to `gztell'
alignment_import.cpp:(.text+0x10b6): undefined reference to `gzrewind'
alignment_import.cpp:(.text+0x10c6): undefined reference to `gzseek'
alignment_import.cpp:(.text+0x10ee): undefined reference to `gzread'
alignment_import.cpp:(.text+0x111e): undefined reference to `gzread'
alignment_import.cpp:(.text+0x1165): undefined reference to `gzrewind'
alignment_import.cpp:(.text+0x1177): undefined reference to `gzseek'
alignment_import.cpp:(.text+0x1228): undefined reference to `gzread'
alignment_import.cpp:(.text+0x12a5): undefined reference to `gzrewind'
alignment_import.cpp:(.text+0x12b7): undefined reference to `gzseek'
alignment_import.cpp:(.text+0x1350): undefined reference to `gzread'
alignment_import.o: In function `loadFromBin(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >)':
alignment_import.cpp:(.text+0x1a17): undefined reference to `gzopen'
alignment_import.cpp:(.text+0x1c91): undefined reference to `gzread'
alignment_import.cpp:(.text+0x1cb6): undefined reference to `gzread'
alignment_import.cpp:(.text+0x1ddc): undefined reference to `gzread'
alignment_import.cpp:(.text+0x1e3c): undefined reference to `gzread'
alignment_import.cpp:(.text+0x2016): undefined reference to `gzread'
alignment_import.o:alignment_import.cpp:(.text+0x2361): more undefined references to `gzread' follow
alignment_import.o: In function `loadFromBin(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >)':
alignment_import.cpp:(.text+0x51cb): undefined reference to `gztell'
alignment_import.cpp:(.text+0x5260): undefined reference to `gzread'
alignment_import.o: In function `fileSeek(gzFile_s*, std::basic_ifstream<char, std::char_traits<char> >&, long)':
alignment_import.cpp:(.text+0xa1a): undefined reference to `gzseek'
alignment_import.o: In function `fileTell(gzFile_s*, std::basic_ifstream<char, std::char_traits<char> >&)':
alignment_import.cpp:(.text+0xa93): undefined reference to `gztell'
collect2: error: ld returned 1 exit status
Makefile:9: recipe for target 'emase-zero' failed
make: *** [emase-zero] Error 1

It seems to work when I move the zlib linker flag (-lz) to the end:

g++ -o emase-zero alignment_incidence_matrix.o sample_allelic_expression.o alignment_import.o emase.o -lz

I found this on the topic: https://stackoverflow.com/questions/9417169/why-does-the-library-linker-flag-sometimes-have-to-go-at-the-end-using-gcc

Thank,

Jonas