[PATCH] Add pbtranscript-tofu.

[Ricardo Wurmus]

Hi Guix,

attached is a patch to add a couple of transcriptome analysis scripts.
I had to use the git downloader because the releases are unusable: they
are really old and only contain the scripts but no longer the setup.py
script.

The git clone also isn't that nice; I had to delete some clutter after
unpacking and fix the permissions on the python scripts or else
installation would fail.

About the name: "pbtranscript(-tofu)" is the correct name, whereas
"cDNA_primer" is not, even though the repository is named "cDNA_primer".

This patch depends on three other packages:

- python2-pbcore:
  http://lists.gnu.org/archive/html/guix-devel/2015-03/msg00573.html

- python2-h5py:
  http://lists.gnu.org/archive/html/guix-devel/2015-03/msg00532.html

- python2-bx-python:
  attached

~~ Ricardo

>From 39f77a513b2505d2a9658f6f9ed40f24a6b93dfd Mon Sep 17 00:00:00 2001
From: Ricardo Wurmus <address@hidden>
Date: Wed, 18 Mar 2015 14:39:42 +0100
Subject: [PATCH 1/2] gnu: Add python2-bx-python.

* gnu/packages/bioinformatics.scm (python2-bx-python): New variable.
---
 gnu/packages/bioinformatics.scm | 35 +++++++++++++++++++++++++++++++++++
 1 file changed, 35 insertions(+)

diff --git a/gnu/packages/bioinformatics.scm b/gnu/packages/bioinformatics.scm
index a8592d0..4ab6054 100644
--- a/gnu/packages/bioinformatics.scm
+++ b/gnu/packages/bioinformatics.scm
@@ -289,6 +289,41 @@ and more accurate.  BWA-MEM also has better performance 
than BWA-backtrack for
 70-100bp Illumina reads.")
     (license license:gpl3+)))
 
+(define-public python2-bx-python
+  (package
+    (name "python2-bx-python")
+    (version "0.7.2")
+    (source (origin
+              (method url-fetch)
+              (uri (string-append
+                    
"https://pypi.python.org/packages/source/b/bx-python/bx-python-";
+                    version ".tar.gz"))
+              (sha256
+               (base32
+                "0ld49idhc5zjdvbhvjq1a2qmpjj7h5v58rqr25dzmfq7g34b50xh"))
+              (modules '((guix build utils)))
+              (snippet
+               '(substitute* "setup.py"
+                  ;; remove dependency on outdated "distribute" module
+                  (("^from distribute_setup import use_setuptools") "")
+                  (("^use_setuptools\\(\\)") "")))))
+    (build-system python-build-system)
+    (arguments
+     `(#:tests? #f ;tests fail because test data are not included
+       #:python ,python-2))
+    (inputs
+     `(("python-numpy" ,python2-numpy)
+       ("zlib" ,zlib)))
+    (native-inputs
+     `(("python-nose" ,python2-nose)
+       ("python-setuptools" ,python2-setuptools)))
+    (home-page "http://bitbucket.org/james_taylor/bx-python/";)
+    (synopsis "Tools for manipulating biological data")
+    (description
+     "bx-python provides tools for manipulating biological data, particularly
+multiple sequence alignments.")
+    (license license:expat)))
+
 (define-public clipper
   (package
     (name "clipper")
-- 
2.1.0

>From 42c865d2c29b7c68e81580de98997266112d6cd9 Mon Sep 17 00:00:00 2001
From: Ricardo Wurmus <address@hidden>
Date: Fri, 20 Mar 2015 14:36:38 +0100
Subject: [PATCH 2/2] gnu: Add pbtranscript-tofu.

* gnu/packages/bioinformatics.scm (pbtranscript-tofu): New variable.
---
 gnu/packages/bioinformatics.scm | 54 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 54 insertions(+)

diff --git a/gnu/packages/bioinformatics.scm b/gnu/packages/bioinformatics.scm
index 4ab6054..46ce4f4 100644
--- a/gnu/packages/bioinformatics.scm
+++ b/gnu/packages/bioinformatics.scm
@@ -20,6 +20,7 @@
   #:use-module ((guix licenses) #:prefix license:)
   #:use-module (guix packages)
   #:use-module (guix download)
+  #:use-module (guix git-download)
   #:use-module (guix build-system gnu)
   #:use-module (guix build-system cmake)
   #:use-module (guix build-system python)
@@ -738,6 +739,59 @@ that a read originated from a particular isoform.")
 files and writing bioinformatics applications.")
     (license license:bsd-3)))
 
+(define-public pbtranscript-tofu
+  (package
+    (name "pbtranscript-tofu")
+    (version "0.4.1-c7bbd5472")
+    (source (origin
+              (method git-fetch)
+              (uri (git-reference
+                    (url 
"https://github.com/PacificBiosciences/cDNA_primer.git";)
+                    (commit "c7bbd54720df578777743d687ceaf6b4f6d4919e")))
+              (file-name (string-append name "-" version ".tar.gz"))
+              (sha256
+               (base32
+                "148xkzi689c49g6fdhckp6mnmj2qhjdf1j4wifm6ja7ij95d7fxx"))))
+    (build-system python-build-system)
+    (arguments
+     `(#:python ,python-2
+       ;; With standard flags, the install phase attempts to create a zip'd
+       ;; egg file, and fails with an error: 'ZIP does not support timestamps
+       ;; before 1980'
+       #:configure-flags '("--single-version-externally-managed"
+                           "--record=pbtranscript-tofu.txt")
+       #:phases
+       (alist-cons-after
+        'unpack 'enter-directory-and-clean-up
+        (lambda _
+          (chdir "pbtranscript-tofu/pbtranscript/")
+          ;; Delete clutter
+          (delete-file-recursively "dist/")
+          (delete-file-recursively "setuptools_cython-0.2.1-py2.6.egg/")
+          (delete-file-recursively "pbtools.pbtranscript.egg-info")
+          (delete-file "Cython-0.20.1.tar.gz")
+          (delete-file "setuptools_cython-0.2.1-py2.7.egg")
+          (delete-file "setuptools_cython-0.2.1.tar.gz")
+          (delete-file "setup.cfg")
+          ;; files should be writable for install phase
+          (for-each (lambda (f) (chmod f #o755))
+                    (find-files "." "\\.py")))
+        %standard-phases)))
+    (inputs
+     `(("python-cython" ,python2-cython)
+       ("python-numpy" ,python2-numpy)
+       ("python-bx-python" ,python2-bx-python)
+       ("python-pbcore" ,python2-pbcore)))
+    (native-inputs
+     `(("python-nose" ,python2-nose)
+       ("python-setuptools" ,python2-setuptools)))
+    (home-page "https://github.com/PacificBiosciences/cDNA_primer";)
+    (synopsis "Analyze transcriptome data generated with the Iso-Seq protocol")
+    (description
+     "pbtranscript-tofu contains scripts to analyze transcriptome data
+generated using the PacBio Iso-Seq protocol.")
+    (license license:bsd-3)))
+
 (define-public rseqc
   (package
     (name "rseqc")
-- 
2.1.0