Merge pull request #11 from bioinform/docker-update

Docker update

docker/Dockerfile

@@ -1,6 +1,6 @@
 FROM ubuntu:16.04
 
-
+ENV R_VERSION 3.3.2
 ENV SAMTOOLS_VERSION 1.2
 ENV BEDTOOLS2_VERSION 2.24.0
 ENV PYBEDTOOLS_VERSION 0.7.7
@@ -19,16 +19,18 @@ ENV BWA_VERSION 0.7.15
 ENV SRA_VERSION 2.8.1
 ENV COREUTILS_VERSION 8.25
 ENV PIGZ_VERSION 2.3.1
+ENV BBMAP_VERSION 37.28
 ENV GMAP_VERSION 2017-02-15
 ENV IDPFUSION_VERSION 1.1.1
 
 RUN apt-get update && \
-    apt-get install -y --fix-missing build-essential zlib1g-dev unzip libncurses5-dev curl wget r-base r-base-dev python python-pip python-dev cmake libboost-all-dev  libxml2-dev libcurl4-gnutls-dev software-properties-common apt-transport-https default-jre default-jdk less vim libtbb-dev
+    apt-get install -y --fix-missing build-essential zlib1g-dev unzip libncurses5-dev curl wget python python-pip python-dev cmake libboost-all-dev  libxml2-dev libcurl4-gnutls-dev software-properties-common apt-transport-https default-jre default-jdk less vim libtbb-dev git
 
 RUN apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv-keys E298A3A825C0D65DFD57CBB651716619E084DAB9
 RUN add-apt-repository 'deb [arch=amd64,i386] https://cran.rstudio.com/bin/linux/ubuntu xenial/'
 RUN apt-get update
-RUN apt-get install -y --fix-missing r-base r-base-dev
+RUN apt-get install -y --fix-missing r-base=${R_VERSION}-1xenial0  r-recommended=${R_VERSION}-1xenial0 
+RUN apt-get install -y --fix-missing --allow-downgrades r-base-core=${R_VERSION}-1xenial0
 
 RUN echo 'local({r <- getOption("repos"); r["CRAN"] <- "http://cran.r-project.org"; options(repos=r)})' > ~/.Rprofile
 RUN R -e 'source("http://bioconductor.org/biocLite.R"); biocLite("DESeq2"); biocLite("tximport"); biocLite("readr");'
@@ -87,16 +89,16 @@ RUN cd /opt && unzip bowtie2-${BOWTIE2_VERSION}-linux-x86_64.zip && cp -p /opt/b
 RUN wget https://sourceforge.net/projects/bio-bwa/files/bwa-${BWA_VERSION}.tar.bz2/download -O /opt/bwa-${BWA_VERSION}.tar.bz2 
 RUN cd /opt && tar xjf bwa-${BWA_VERSION}.tar.bz2 && cd bwa-${BWA_VERSION} && make && cp -p /opt/bwa-${BWA_VERSION}/bwa /usr/local/bin && cd /opt && rm -rf bwa*
 
-ADD https://github.com/ndaniel/seqtk/archive/1.0-r82b.tar.gz /opt/seqtk-1.0-r82b.tar.gz
-RUN cd /opt && tar -zxvf /opt/seqtk-1.0-r82b.tar.gz && cd seqtk-1.0-r82b && make  && cp -p /opt/seqtk-1.0-r82b/seqtk /usr/local/bin && cd /opt && rm -rf seqtk*
+ADD https://github.com/ndaniel/seqtk/archive/1.2-r101c.tar.gz /opt/seqtk-1.2-r101c.tar.gz
+RUN cd /opt && tar -zxvf /opt/seqtk-1.2-r101c.tar.gz && cd seqtk-1.2-r101c && make  && cp -p /opt/seqtk-1.2-r101c/seqtk /usr/local/bin && cd /opt && rm -rf seqtk*
 
 ADD http://hgdownload.cse.ucsc.edu/admin/exe/linux.x86_64/blat/blat /usr/local/bin/blat
 RUN chmod 755 /usr/local/bin/blat
 
 ADD http://hgdownload.cse.ucsc.edu/admin/exe/linux.x86_64/faToTwoBit /usr/local/bin/faToTwoBit
 RUN chmod 755 /usr/local/bin/faToTwoBit
 
-ADD http://hgdownload.cse.ucsc.edu/admin/exe/linux.x86_64.v287/liftOver /usr/local/bin/liftOver
+ADD http://hgdownload.cse.ucsc.edu/admin/exe/linux.x86_64/liftOver /usr/local/bin/liftOver
 RUN chmod 755 /usr/local/bin/liftOver
 
 ADD https://ftp-trace.ncbi.nlm.nih.gov/sra/sdk/${SRA_VERSION}/sratoolkit.${SRA_VERSION}-ubuntu64.tar.gz /opt/sratoolkit.${SRA_VERSION}-ubuntu64.tar.gz
@@ -105,25 +107,28 @@ RUN cd /opt && tar -zxvf sratoolkit.${SRA_VERSION}-ubuntu64.tar.gz && cp -Rp /op
 ADD http://ftp.gnu.org/gnu/coreutils/coreutils-${COREUTILS_VERSION}.tar.xz /opt/coreutils-${COREUTILS_VERSION}.tar.xz 
 RUN cd /opt && tar -xJf coreutils-${COREUTILS_VERSION}.tar.xz && cd coreutils-${COREUTILS_VERSION} && ./configure FORCE_UNSAFE_CONFIGURE=1 && make && make install && cd /opt && rm -rf coreutils*
 
-ADD http://pkgs.fedoraproject.org/repo/pkgs/pigz/pigz-${PIGZ_VERSION}.tar.gz/e803f8bc0770c7a5e96dccb1d2dd2aab/pigz-${PIGZ_VERSION}.tar.gz /opt/pigz-${PIGZ_VERSION}.tar.gz
+ADD https://github.com/madler/pigz/archive/v${PIGZ_VERSION}.tar.gz /opt/pigz-${PIGZ_VERSION}.tar.gz
 RUN cd /opt && tar -zxvf pigz-${PIGZ_VERSION}.tar.gz && cd pigz-${PIGZ_VERSION} && make && cp -p /opt/pigz-${PIGZ_VERSION}/pigz /usr/local/bin && cd /opt && rm -rf pigz*
 
 ADD http://research-pub.gene.com/gmap/src/gmap-gsnap-${GMAP_VERSION}.tar.gz /opt/gmap-gsnap-${GMAP_VERSION}.tar.gz
 RUN cd /opt && tar -zxvf gmap-gsnap-${GMAP_VERSION}.tar.gz && cd gmap-${GMAP_VERSION} && ./configure && make && make install && cd /opt && rm -rf gmap*
 
-RUN wget https://github.com/ndaniel/fusioncatcher/archive/master.zip -O /opt/fusioncatcher-master.zip && cd /opt && unzip fusioncatcher-master.zip  && cp -p /opt/fusioncatcher-master/bin/sam2psl.py /usr/local/bin
-ENV PATH $PATH:/opt/fusioncatcher-master/bin/
+RUN wget https://sourceforge.net/projects/bbmap/files/BBMap_${BBMAP_VERSION}.tar.gz -O /opt/BBMap_${BBMAP_VERSION}.tar.gz
+RUN cd /opt && tar -xzvf BBMap_${BBMAP_VERSION}.tar.gz
+ENV PATH $PATH:/opt/bbmap/
+
+RUN cd /opt/ && git clone https://github.com/ndaniel/fusioncatcher.git && cd fusioncatcher && git checkout 60bddd2f1ddfd95a2dbc6c4efa3c521bea3421da &&  cp -p /opt/fusioncatcher/bin/sam2psl.py /usr/local/bin
+ENV PATH $PATH:/opt/fusioncatcher/bin/
+
+ADD http://ccb.jhu.edu/software/stringtie/dl/gffread-0.9.12.Linux_x86_64.tar.gz opt/gffread-0.9.12.Linux_x86_64.tar.gz 
+RUN cd /opt && tar -xzvf gffread-0.9.12.Linux_x86_64.tar.gz && cp -p /opt/gffread-0.9.12.Linux_x86_64/gffread /usr/local/bin && rm -rf /opt/gffread*
 
-ADD https://github.com/bioinform/IDP/archive/master.zip /opt/IDP-master.zip
-RUN rm -rf /usr/local/bin/IDP* && cd /opt && unzip IDP-master.zip && rm -rf /opt/IDP-master.zip
+RUN cd /opt/ && git clone https://github.com/bioinform/IDP.git && cd IDP && git checkout a5d2d624ab8e4545feff3f51d264931b440d0b53
 
 ADD https://www.healthcare.uiowa.edu/labs/au/IDP-fusion/files/IDP-fusion_${IDPFUSION_VERSION}.tar.gz /opt/IDP-fusion_${IDPFUSION_VERSION}.tar.gz
 RUN cd /opt && tar -xzvf IDP-fusion_${IDPFUSION_VERSION}.tar.gz && rm -rf /opt/IDP-fusion_${IDPFUSION_VERSION}.tar.gz
 
-
-RUN pip install https://github.com/bioinform/RNACocktail/archive/v0.2.tar.gz
+RUN pip install https://github.com/bioinform/rnacocktail/archive/v0.2.2.tar.gz
 
 VOLUME /work_dir
 
-
-

index.html

@@ -150,17 +150,19 @@ <h2>A comprehensive framework for accurate and efficient RNA-Seq analysis</h2>
 
         </div>
 
-    <h2>Publication [Open access]</h2>
+    <h2>Publication</h2>
 
     <div class="panel panel-default" style="font-family:monospace;">
         <div class="panel-body">
-<!-- 
             <i>If you use RNACocktail in your work, please cite the following:</i><br>
-            ??<br>
-            <b>??</b><br>
-            ??
- -->
- 			Manuscript in preparation.
+            Sayed Mohammad Ebrahim Sahraeian, Marghoob Mohiyuddin, Robert Sebra, Hagen Tilgner, 
+            Pegah T. Afshar, Kin Fai Au, Narges Bani Asadi, Mark B. Gerstein, Wing Hung Wong, 
+            Michael P. Snyder, Eric Schadt, and Hugo Y. K. Lam<br>
+            <b>Gaining comprehensive biological insight into the transcriptome by performing a broad-spectrum RNA-seq analysis</b><br>
+            Nature Communications 8, Article number: 59 (2017). <a
+                href="http://dx.doi.org/10.1038/s41467-017-00050-4"
+                onclick="trackOutboundLink('http://dx.doi.org/10.1038/s41467-017-00050-4'); return false;">doi:10.1038/s41467-017-00050-4
+</a>
         </div>
     </div>
 

src/_version.py

@@ -1 +1 @@
-__version__ = "0.2.1"
+__version__ = "0.2.2"

src/run_diff.py

@@ -31,7 +31,17 @@ def tx2gene_map(ref_gtf_file,tx2gene_file):
         spamwriter.writerows(tx2gene.items())
     return tx2gene
 
-
+def fix_quant_file(quant_file,fixed_quant_file):
+    cnt=0
+    with open(quant_file) as q_f:
+        with open(fixed_quant_file,"w'") as fixed_q_f:
+            for line in q_f:
+                if cnt>0:
+                    fields=line.strip().split("\t")
+                    fields[0]=fields[0].split("|")[0]
+                    line="\t".join(fields)+"\n"
+                fixed_q_f.write(line)
+                cnt+=1
 
 def run_deseq2(quant_files="", alignments="",
               transcripts_gtfs="", ref_gtf="", 
@@ -140,6 +150,16 @@ def run_deseq2(quant_files="", alignments="",
         msg="compute gene level abundances for %s."%samples_txt
         if start<=step:
             logger.info("--------------------------STEP %s--------------------------"%step)
+
+            fixed_quant_files=[]
+            for i,qs in enumerate(quant_files):
+                fixed_qs=[]
+                for j,q in enumerate(qs):
+                    fixed_q =  os.path.join(work_deseq2, "{}.fixed_quant.sf".format(samples[i][j]))
+                    fix_quant_file(q,fixed_q)
+                    fixed_qs.append(fixed_q)
+                fixed_quant_files.append(fixed_qs)
+
             command = "%s -e \"library('readr'); library('tximport'); \
                        samples=c(%s); (files <- file.path(c(%s))); names(files) <- samples; \
                        tx2gene <- read.csv(file.path('%s'),sep='\\t'); \
@@ -150,7 +170,7 @@ def run_deseq2(quant_files="", alignments="",
                        write.table(txi$length, file = '%s/txi.length', quote = FALSE, \
                        sep='\\t'); write.table(txi$counts, file = '%s/txi.counts',\
                        quote = FALSE, sep='\\t');\""%(R, ",".join(map(lambda x: "'%s'"%x,reduce(lambda x,y:x+y,samples)))
-                                              ,",".join(map(lambda x: "'%s'"%x,reduce(lambda x,y:x+y,quant_files)))
+                                              ,",".join(map(lambda x: "'%s'"%x,reduce(lambda x,y:x+y,fixed_quant_files)))
                                               ,tx2gene_file, work_deseq2, work_deseq2, work_deseq2, work_deseq2)
             cmd = TimedExternalCmd(command, logger, raise_exception=True)
             retcode = cmd.run(cmd_log_fd_out=deseq2_log_fd, cmd_log=deseq2_log, msg=msg, timeout=timeout)

src/run_lr_correct.py

@@ -43,8 +43,8 @@ def run_lordec(kmer=23,
     lordec_log_fd = open(lordec_log, "w")
     ksps = ""
 
-    if "--threads " not in lordec_opts:
-        lordec_opts += " --T %d"%nthreads 
+    if "-T " not in lordec_opts:
+        lordec_opts += " -T %d"%nthreads 
 
 
     msg = "LoRDEC for %s"%sample

src/run_lr_fusion.py

@@ -22,7 +22,7 @@ def sort_gpd(in_file,out_file,order_chrs=dict([("%s"%k,k) for k in range(1,23)]+
         others_chrs=sorted(set(map(lambda x:x[2],rows))-set(order_chrs.keys()))
         if others_chrs:
             max_id=max(order_chrs.values())
-            for i,c in enumerated(others_chrs):
+            for i,c in enumerate(others_chrs):
                 order_chrs[c]=max_id+i+1
         sorted_rows=sorted(rows,key=lambda x: (order_chrs[x[2]],int(x[4])))
         with open(out_file, 'wb') as csvfile:

src/run_lr_reconstruct.py

@@ -21,7 +21,7 @@ def sort_gpd(in_file,out_file,order_chrs=dict([("%s"%k,k) for k in range(1,23)]+
         others_chrs=sorted(set(map(lambda x:x[2],rows))-set(order_chrs.keys()))
         if others_chrs:
             max_id=max(order_chrs.values())
-            for i,c in enumerated(others_chrs):
+            for i,c in enumerate(others_chrs):
                 order_chrs[c]=max_id+i+1
         sorted_rows=sorted(rows,key=lambda x: (order_chrs[x[2]],int(x[4])))
         with open(out_file, 'wb') as csvfile: