replace local AnnoSINEv2 with the conda version

oushujun · Jan 12, 2024 · 8980f49 · 8980f49
1 parent 34b6d40
commit 8980f49
Show file tree

Hide file tree

Showing 3 changed files with 83 additions and 66 deletions.
diff --git a/EDTA.pl b/EDTA.pl
@@ -269,11 +269,11 @@
 $genometools="$genometools/" if $genometools ne '' and $genometools !~ /\/$/;
 die "Error: gt is not found in the genometools path $genometools!\n" unless -X "${genometools}gt";
 # AnnoSINE
-chomp ($annosine=`which annosine2 2>/dev/null`) if $annosine eq '';
+chomp ($annosine=`which AnnoSINE_v2 2>/dev/null`) if $annosine eq '';
 $annosine =~ s/\s+$//;
 $annosine = dirname($annosine) unless -d $annosine;
 $annosine="$annosine/" if $annosine ne '' and $annosine !~ /\/$/;
-die "Error: AnnoSINE is not found in the AnnoSINE path $annosine!\n" unless (-X "${annosine}AnnoSINE_v2.py" or -X "${annosine}/bin/AnnoSINE_v2.py" or -X "${annosine}annosine2");
+die "Error: AnnoSINE is not found in the AnnoSINE path $annosine!\n" unless (-X "${annosine}AnnoSINE_v2");
 # LTR_retriever
 chomp ($LTR_retriever=`which LTR_retriever 2>/dev/null`) if $LTR_retriever eq '';
 $LTR_retriever =~ s/\s+$//;

diff --git a/EDTA_2.2.x.yml b/EDTA_2.2.x.yml
@@ -1,4 +1,4 @@
-name: EDTA2.2
+name: EDTA2
 channels:
   - conda-forge
   - bioconda
@@ -12,35 +12,24 @@ dependencies:
   - aiohttp=3.9.1=py39hd1e30aa_0
   - aiosignal=1.3.1=pyhd8ed1ab_0
   - alsa-lib=1.2.9=hd590300_0
-  - arrow-cpp=10.0.1=h3e2b116_4_cpu
+  - annosine2=1.0.5=pyh7cba7a3_0
   - astor=0.8.1=pyh9f0ad1d_0
   - astunparse=1.6.3=pyhd8ed1ab_0
   - async-timeout=4.0.3=pyhd8ed1ab_0
   - attrs=23.2.0=pyh71513ae_0
-  - aws-c-auth=0.6.21=hd93a3ba_3
-  - aws-c-cal=0.5.20=hd3b2fe5_3
-  - aws-c-common=0.8.5=h166bdaf_0
-  - aws-c-compression=0.2.16=hf5f93bc_0
-  - aws-c-event-stream=0.2.17=h57874a7_1
-  - aws-c-http=0.7.0=h96ef541_0
-  - aws-c-io=0.13.12=h57ca295_1
-  - aws-c-mqtt=0.7.13=h0b5698f_12
-  - aws-c-s3=0.2.1=h5d7e488_4
-  - aws-c-sdkutils=0.1.7=hf5f93bc_0
-  - aws-checksums=0.1.14=h6027aba_0
-  - aws-crt-cpp=0.18.16=h26430d7_8
-  - aws-sdk-cpp=1.9.379=h56135f1_7
   - bedtools=2.31.1=hf5e1c6e_0
   - binutils_impl_linux-64=2.40=hf600244_0
-  - biopython=1.82=py39hd1e30aa_0
+  - biopython=1.83=py39hd1e30aa_0
   - blast=2.10.1=pl526he19e7b1_3
   - blinker=1.7.0=pyhd8ed1ab_0
-  - bokeh=3.3.2=pyhd8ed1ab_0
-  - brotli-python=1.0.9=py39h5a03fae_9
+  - bokeh=2.4.3=pyhd8ed1ab_3
+  - brotli=1.1.0=hd590300_1
+  - brotli-bin=1.1.0=hd590300_1
+  - brotli-python=1.1.0=py39h3d6467e_1
   - bwidget=1.9.14=ha770c72_1
   - bz2file=0.98=py_0
   - bzip2=1.0.8=hd590300_5
-  - c-ares=1.24.0=hd590300_0
+  - c-ares=1.25.0=hd590300_0
   - ca-certificates=2023.11.17=hbcca054_0
   - cached-property=1.5.2=hd8ed1ab_1
   - cached_property=1.5.2=pyha770c72_1
@@ -62,10 +51,10 @@ dependencies:
   - curl=7.88.1=h37d81fd_2
   - cycler=0.12.1=pyhd8ed1ab_0
   - cytoolz=0.12.2=py39hd1e30aa_1
-  - dask=2023.12.1=pyhd8ed1ab_0
-  - dask-core=2023.12.1=pyhd8ed1ab_0
+  - dask=2023.3.0=pyhd8ed1ab_0
+  - dask-core=2023.3.0=pyhd8ed1ab_0
   - dill=0.3.7=pyhd8ed1ab_0
-  - distributed=2023.12.1=pyhd8ed1ab_0
+  - distributed=2023.3.0=pyhd8ed1ab_0
   - drmaa=0.7.9=py_1000
   - entrez-direct=16.2=he881be0_1
   - expat=2.5.0=hcb278e6_1
@@ -77,40 +66,40 @@ dependencies:
   - fontconfig=2.14.2=h14ed4e7_0
   - fonts-conda-ecosystem=1=0
   - fonts-conda-forge=1=0
+  - fonttools=4.47.2=py39hd1e30aa_0
   - freetype=2.12.1=h267a509_2
   - fribidi=1.0.10=h36c2ea0_0
   - frozenlist=1.4.1=py39hd1e30aa_0
   - fsspec=2023.12.2=pyhca7485f_0
   - gast=0.4.0=pyh9f0ad1d_0
   - gawk=5.1.0=h7f98852_0
-  - gcc_impl_linux-64=12.3.0=he2b93b0_3
+  - gcc_impl_linux-64=13.2.0=h338b0a0_3
   - genericrepeatfinder=1.0.2=h4ac6f70_0
   - genometools-genometools=1.6.4=py39h58cc16e_0
   - gettext=0.21.1=h27087fc_0
-  - gflags=2.2.2=he1b5a44_1004
-  - gfortran_impl_linux-64=12.3.0=hfcedea8_3
+  - gfortran_impl_linux-64=13.2.0=h76e1118_3
   - giflib=5.2.1=h0b41bf4_3
   - glob2=0.7=py_0
-  - glog=0.6.0=h6f12383_0
   - gmp=6.1.2=hf484d3e_1000
   - gnutls=3.5.19=h2a4e5f8_1
-  - google-auth=2.25.2=pyhca7485f_0
+  - google-auth=2.26.2=pyhca7485f_0
   - google-auth-oauthlib=0.4.6=pyhd8ed1ab_0
   - google-pasta=0.2.0=pyh8c360ce_0
   - graphite2=1.3.13=h58526e2_1001
   - grpcio=1.51.1=py39h712372c_0
   - gsl=2.7=he838d99_0
-  - gxx_impl_linux-64=12.3.0=he2b93b0_3
+  - gxx_impl_linux-64=13.2.0=h338b0a0_3
   - h5py=3.9.0=nompi_py39h4dfffb9_100
   - harfbuzz=6.0.0=h8e241bc_0
   - hdf5=1.14.0=nompi_h5231ba7_103
   - hmmer=3.4=hdbdd923_0
   - icu=70.1=h27087fc_0
   - idna=3.6=pyhd8ed1ab_0
   - importlib-metadata=7.0.1=pyha770c72_0
-  - importlib_metadata=7.0.1=hd8ed1ab_0
-  - irfinder=1.3.1=h031d066_4
-  - jinja2=3.1.2=pyhd8ed1ab_1
+  - importlib-resources=6.1.1=pyhd8ed1ab_0
+  - importlib_resources=6.1.1=pyhd8ed1ab_0
+  - irf=3.08=hdbdd923_0
+  - jinja2=3.1.3=pyhd8ed1ab_0
   - joblib=1.3.2=pyhd8ed1ab_0
   - jpeg=9e=h166bdaf_2
   - k8=0.2.5=hdcf5f25_4
@@ -125,30 +114,26 @@ dependencies:
   - lerc=4.0.0=h27087fc_0
   - libabseil=20220623.0=cxx17_h05df665_6
   - libaec=1.1.2=h59595ed_1
-  - libarrow=10.0.1=hee49ebd_4_cpu
   - libblas=3.9.0=20_linux64_openblas
-  - libbrotlicommon=1.0.9=h166bdaf_9
-  - libbrotlidec=1.0.9=h166bdaf_9
-  - libbrotlienc=1.0.9=h166bdaf_9
+  - libbrotlicommon=1.1.0=hd590300_1
+  - libbrotlidec=1.1.0=hd590300_1
+  - libbrotlienc=1.1.0=hd590300_1
   - libcblas=3.9.0=20_linux64_openblas
-  - libcrc32c=1.1.2=h9c3ff4c_0
   - libcups=2.3.3=h36d4200_3
   - libcurl=7.88.1=h91b91d3_2
   - libdeflate=1.17=h0b41bf4_0
   - libedit=3.1.20191231=he28a2e2_2
   - libev=4.33=hd590300_2
-  - libevent=2.1.10=h9b69904_4
   - libexpat=2.5.0=hcb278e6_1
   - libffi=3.4.2=h7f98852_5
   - libgcc=7.2.0=h69d50b8_2
-  - libgcc-devel_linux-64=12.3.0=h8bca6fd_103
+  - libgcc-devel_linux-64=13.2.0=ha9c7c90_103
   - libgcc-ng=13.2.0=h807b86a_3
   - libgd=2.3.3=h5aea950_4
   - libgfortran-ng=13.2.0=h69a702a_3
   - libgfortran5=13.2.0=ha4646dd_3
   - libglib=2.78.1=hebfc3b9_0
   - libgomp=13.2.0=h807b86a_3
-  - libgoogle-cloud=2.5.0=h5d50b59_1
   - libgrpc=1.51.1=h05bd8bd_0
   - libiconv=1.17=hd590300_2
   - libidn2=2.3.4=h166bdaf_0
@@ -159,15 +144,13 @@ dependencies:
   - libopenblas=0.3.25=pthreads_h413a1c8_0
   - libpng=1.6.39=h753d276_0
   - libprotobuf=3.21.12=hfc55251_2
-  - libsanitizer=12.3.0=h0f45ef3_3
+  - libsanitizer=13.2.0=h7e041cc_3
   - libsqlite=3.44.2=h2797004_0
   - libssh2=1.10.0=haa6b8db_3
-  - libstdcxx-devel_linux-64=12.3.0=h8bca6fd_103
-  - libstdcxx-ng=12.3.0=h0f45ef3_3
-  - libthrift=0.16.0=h491838f_2
+  - libstdcxx-devel_linux-64=13.2.0=ha9c7c90_103
+  - libstdcxx-ng=13.2.0=h7e041cc_3
   - libtiff=4.5.0=h6adf6a1_2
   - libunistring=0.9.10=h7f98852_0
-  - libutf8proc=2.8.0=h166bdaf_0
   - libuuid=2.38.1=h0b41bf4_0
   - libwebp=1.2.4=h1daa5a0_1
   - libwebp-base=1.2.4=h166bdaf_0
@@ -176,38 +159,36 @@ dependencies:
   - libzlib=1.2.13=hd590300_5
   - locket=1.0.0=pyhd8ed1ab_0
   - ltr_finder=1.07=h4ac6f70_3
-  - ltr_retriever=2.9.5=hdfd78af_0
-  - lz4=4.3.2=py39h79d96da_1
+  - ltr_retriever=2.9.9=hdfd78af_0
+  - lz4=4.3.3=py39h79d96da_0
   - lz4-c=1.9.4=hcb278e6_0
   - mafft=7.520=h031d066_3
   - make=4.3=hd18ef5c_1
-  - markdown=3.5.1=pyhd8ed1ab_0
+  - markdown=3.5.2=pyhd8ed1ab_0
   - markupsafe=2.1.3=py39hd1e30aa_1
-  - matplotlib=3.3.2=0
-  - matplotlib-base=3.3.2=py39h98787fa_1
+  - matplotlib-base=3.8.2=py39he9076e7_0
   - mdust=2006.10.17=h031d066_6
   - minimap2=2.26=he4a0461_2
   - msgpack-python=1.0.7=py39h7633fee_0
   - multidict=6.0.4=py39hd1e30aa_1
   - multiprocess=0.70.15=py39hd1e30aa_1
+  - munkres=1.0.7=py_1
   - muscle=5.1=h4ac6f70_3
   - mysql-connector-c=6.1.11=h6eb9d5d_1007
   - nccl=2.19.4.1=h6103f9b_0
   - ncurses=6.4=h59595ed_2
   - nettle=3.3=0
   - nodejs=12.4.0=he1b5a44_0
   - nseg=1.0.1=h031d066_4
-  - numpy=1.26.2=py39h474f0d3_0
+  - numpy=1.26.3=py39h474f0d3_0
   - oauthlib=3.2.2=pyhd8ed1ab_0
   - openjdk=11.0.1=h516909a_1016
   - openjpeg=2.5.0=hfec8fc6_2
   - openssl=1.1.1w=hd590300_0
   - opt_einsum=3.3.0=pyhc1e730c_2
-  - orc=1.8.2=hfdbbad2_2
   - packaging=23.2=pyhd8ed1ab_0
   - pandas=2.1.4=py39hddac248_0
   - pango=1.50.14=hd33c08f_0
-  - parquet-cpp=1.5.1=2
   - partd=1.4.1=pyhd8ed1ab_0
   - pcre=8.45=h9c3ff4c_0
   - pcre2=10.40=hc3806b6_0
@@ -243,13 +224,11 @@ dependencies:
   - perl-xsloader=0.24=pl526_0
   - pillow=9.4.0=py39h2320bf1_1
   - pip=23.3.2=pyhd8ed1ab_0
-  - pixman=0.42.2=h59595ed_0
+  - pixman=0.43.0=h59595ed_0
   - pp=1.6.4.4=py_0
   - protobuf=4.21.12=py39h227be39_0
   - psutil=5.9.7=py39hd1e30aa_0
   - pthread-stubs=0.4=h36c2ea0_1001
-  - pyarrow=10.0.1=py39hacc6ce7_4_cpu
-  - pyarrow-hotfix=0.6=pyhd8ed1ab_0
   - pyasn1=0.5.1=pyhd8ed1ab_0
   - pyasn1-modules=0.3.0=pyhd8ed1ab_0
   - pycparser=2.21=pyhd8ed1ab_0
@@ -266,6 +245,45 @@ dependencies:
   - pyu2f=0.1.5=pyhd8ed1ab_0
   - pyyaml=6.0.1=py39hd1e30aa_1
   - r-base=4.2.3=ha7d60f8_0
+  - r-cli=3.6.2=r42ha503ecb_0
+  - r-colorspace=2.1_0=r42h57805ef_1
+  - r-crayon=1.5.2=r42hc72bb7e_2
+  - r-dplyr=1.1.4=r42ha503ecb_0
+  - r-ellipsis=0.3.2=r42h57805ef_2
+  - r-fansi=1.0.6=r42h57805ef_0
+  - r-farver=2.1.1=r42ha503ecb_2
+  - r-generics=0.1.3=r42hc72bb7e_2
+  - r-ggplot2=3.4.4=r42hc72bb7e_0
+  - r-glue=1.7.0=r42h57805ef_0
+  - r-gtable=0.3.4=r42hc72bb7e_0
+  - r-here=1.0.1=r42hc72bb7e_2
+  - r-isoband=0.2.7=r42ha503ecb_2
+  - r-labeling=0.4.3=r42hc72bb7e_0
+  - r-lattice=0.22_5=r42h57805ef_0
+  - r-lifecycle=1.0.4=r42hc72bb7e_0
+  - r-magrittr=2.0.3=r42h57805ef_2
+  - r-mass=7.3_60=r42h57805ef_1
+  - r-matrix=1.6_5=r42h316c678_0
+  - r-mgcv=1.9_1=r42h316c678_0
+  - r-munsell=0.5.0=r42hc72bb7e_1006
+  - r-nlme=3.1_164=r42h61816a4_0
+  - r-pillar=1.9.0=r42hc72bb7e_1
+  - r-pkgconfig=2.0.3=r42hc72bb7e_3
+  - r-purrr=1.0.2=r42h57805ef_0
+  - r-r6=2.5.1=r42hc72bb7e_2
+  - r-rcolorbrewer=1.1_3=r42h785f33e_2
+  - r-rlang=1.1.3=r42ha503ecb_0
+  - r-rprojroot=2.0.4=r42hc72bb7e_0
+  - r-scales=1.3.0=r42hc72bb7e_0
+  - r-stringi=1.7.12=r42h1ae9187_0
+  - r-stringr=1.5.1=r42h785f33e_0
+  - r-tibble=3.2.1=r42h57805ef_2
+  - r-tidyr=1.3.0=r42ha503ecb_1
+  - r-tidyselect=1.2.0=r42hc72bb7e_1
+  - r-utf8=1.2.4=r42h57805ef_0
+  - r-vctrs=0.6.5=r42ha503ecb_0
+  - r-viridislite=0.4.2=r42hc72bb7e_1
+  - r-withr=2.5.2=r42hc72bb7e_0
   - re2=2022.06.01=h27087fc_1
   - readline=8.2=h8228510_1
   - recon=1.08=h031d066_6
@@ -277,12 +295,11 @@ dependencies:
   - requests-oauthlib=1.3.1=pyhd8ed1ab_0
   - rmblast=2.10.0=h2d02072_0
   - rsa=4.9=pyhd8ed1ab_0
-  - s2n=1.3.31=hae46d1a_0
   - scikit-learn=1.3.2=py39ha22ef79_2
   - scipy=1.11.4=py39h474f0d3_0
   - sed=4.8=he412f7d_0
   - seqtk=1.4=he4a0461_1
-  - setuptools=68.2.2=pyhd8ed1ab_0
+  - setuptools=69.0.3=pyhd8ed1ab_0
   - six=1.16.0=pyh6c4a22f_0
   - snappy=1.1.10=h9fff704_0
   - sortedcontainers=2.4.0=pyhd8ed1ab_0
@@ -296,7 +313,7 @@ dependencies:
   - tensorflow=2.11.0=cuda112py39h01bd6f0_0
   - tensorflow-base=2.11.0=cuda112py39h1c230a5_0
   - tensorflow-estimator=2.11.0=cuda112py39hd320b7a_0
-  - termcolor=2.3.0=pyhd8ed1ab_0
+  - termcolor=2.4.0=pyhd8ed1ab_0
   - tesorter=1.4.6=pyhdfd78af_0
   - threadpoolctl=3.2.0=pyha21a80b_0
   - tk=8.6.13=noxft_h4845f30_101
@@ -311,6 +328,7 @@ dependencies:
   - ucsc-fatotwobit=447=h954228d_0
   - ucsc-twobitinfo=447=h954228d_0
   - ucsc-twobittofa=447=h954228d_0
+  - unicodedata2=15.1.0=py39hd1e30aa_0
   - urllib3=2.1.0=pyhd8ed1ab_0
   - werkzeug=3.0.1=pyhd8ed1ab_0
   - wget=1.20.3=ha56f1ee_1
@@ -335,7 +353,6 @@ dependencies:
   - xorg-renderproto=0.11.1=h7f98852_1002
   - xorg-xextproto=7.3.0=h0b41bf4_1003
   - xorg-xproto=7.0.31=h7f98852_1007
-  - xyzservices=2023.10.1=pyhd8ed1ab_0
   - xz=5.2.6=h166bdaf_0
   - yaml=0.2.5=h7f98852_2
   - yarl=1.9.3=py39hd1e30aa_0

diff --git a/EDTA_raw.pl b/EDTA_raw.pl
@@ -48,7 +48,6 @@
 	--mdust		[path]	Path to the mdust program. (default: find from ENV)
 	--repeatmasker	[path]	Path to the RepeatMasker program. (default: find from ENV)
 	--repeatmodeler	[path]	Path to the RepeatModeler2 program. (default: find from ENV)
-	--annosine	[path]  The directory containing AnnoSINE (default: read from ENV)
 	--threads|-t	[int]	Number of theads to run this script. Default: 4
 	--help|-h	Display this help info
 \n";
@@ -195,11 +194,11 @@
 $repeatmodeler="$repeatmodeler/" if $repeatmodeler ne '' and $repeatmodeler !~ /\/$/;
 die "Error: RepeatModeler is not found in the RepeatModeler path $repeatmodeler!\n" unless -X "${repeatmodeler}RepeatModeler";
 # AnnoSINE
-chomp ($annosine=`which annosine2 2>/dev/null`) if $annosine eq '';
+chomp ($annosine=`which AnnoSINE_v2 2>/dev/null`) if $annosine eq '';
 $annosine =~ s/\s+$//;
 $annosine = dirname($annosine) unless -d $annosine;
 $annosine="$annosine/" if $annosine ne '' and $annosine !~ /\/$/;
-die "Error: AnnoSINE is not found in the AnnoSINE path $annosine!\n" unless (-X "${annosine}AnnoSINE_v2.py" or -X "${annosine}/bin/AnnoSINE_v2.py" or -X "${annosine}annosine2");
+die "Error: AnnoSINE is not found in the AnnoSINE path $annosine!\n" unless (-X "${annosine}AnnoSINE_v2");
 # LTR_retriever
 chomp ($LTR_retriever=`which LTR_retriever 2>/dev/null`) if $LTR_retriever eq '';
 $LTR_retriever =~ s/\s+$//;
@@ -426,7 +425,7 @@
 if (-s "Seed_SINE.fa"){
 	print STDERR "$date\tExisting result file Seed_SINE.fa found!\n\t\t\t\tWill keep this file without rerunning this module.\n\t\t\t\tPlease specify --overwrite 1 if you want to rerun AnnoSINE_v2.\n\n";
 	} else { 
-	$status = system("python3 ${annosine}annosine2 -t $threads -a 2 --num_alignments 50000 -rpm 0 --copy_number 3 --shift 100 -auto 1 3 $genome ./ > /dev/null 2>&1");
+	$status = system("python3 ${annosine}AnnoSINE_v2 -t $threads -a 2 --num_alignments 50000 -rpm 0 --copy_number 3 --shift 100 -auto 1 3 $genome ./ > /dev/null 2>&1");
 	#`rm $_` for grep { /^.\/${genome}_([0-9a-f]{32})\.mod$/i } glob("./*"); # remove duplicated genome file
 	}
 
@@ -573,7 +572,8 @@
 	if ($overwrite eq 0 and -s "./TIR-Learner-Result/TIR-Learner_FinalAnn.fa"){
 		print STDERR "$date\tExisting raw result TIR-Learner_FinalAnn.fa found!\n\t\t\t\tWill use this for further analyses.\n\t\t\t\tPlease specify --overwrite 1 if you want to rerun this module.\n\n";
 		} else {
-		`python3 $TIR_Learner/TIR-Learner3.0.py -f $genome_file_real_path -s $species -t $threads -l $maxint -c -o $genome_file_real_path.EDTA.raw/TIR --grf_path $grfp --gt_path $genometools`;
+		#`python3 $TIR_Learner/TIR-Learner3.0.py -f $genome_file_real_path -s $species -t $threads -l $maxint -c -o $genome_file_real_path.EDTA.raw/TIR --grf_path $grfp --gt_path $genometools`;
+		`python3 $TIR_Learner/TIR-Learner3.0.py -f $genome_file_real_path -s $species -t $threads -l $maxint -o $genome_file_real_path.EDTA.raw/TIR --grf_path $grfp --gt_path $genometools`;
 		}
 
 	# clean raw predictions with flanking alignment