[PATCH] Add genome annotations.

[Ricardo Wurmus]

Hi Guix,

here are four patches to add genome annotation packages for R.  The
species names (“Mouse”, “Worm”, “Human”, “Fly”) are capitalised because
they refer to the names of common model organisms.

~~ Ricardo

PS: for the curious: “hs” = “Homo sapiens” (human), “ce” =
“Caenorhabditis elegans” (worm), “dm” = “Drosophila melanogaster” (fly),
“mm” = “Mus musculus” (mouse).

>From f0ed886cb4399d8acd15e6fa894042f60a0a4bcd Mon Sep 17 00:00:00 2001
From: Ricardo Wurmus <address@hidden>
Date: Wed, 13 Jan 2016 16:27:29 +0100
Subject: [PATCH 1/4] gnu: Add r-org-hs-eg-db.

* gnu/packages/bioinformatics.scm (r-org-hs-eg-db): New variable.
---
 gnu/packages/bioinformatics.scm | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)

diff --git a/gnu/packages/bioinformatics.scm b/gnu/packages/bioinformatics.scm
index 9b5bd12..fe82329 100644
--- a/gnu/packages/bioinformatics.scm
+++ b/gnu/packages/bioinformatics.scm
@@ -4330,6 +4330,32 @@ genomic feature data as long as it has minimal 
information on the locations of
 genomic intervals.  In addition, it can use BAM or BigWig files as input.")
     (license license:artistic2.0)))
 
+(define-public r-org-hs-eg-db
+  (package
+    (name "r-org-hs-eg-db")
+    (version "3.2.3")
+    (source (origin
+              (method url-fetch)
+              ;; We cannot use bioconductor-uri here because this tarball is
+              ;; located under "data/annotation/" instead of "bioc/".
+              (uri (string-append "http://www.bioconductor.org/packages/";
+                                  "release/data/annotation/src/contrib/"
+                                  "org.Hs.eg.db_" version ".tar.gz"))
+              (sha256
+               (base32
+                "0xicgkbh6xkvs74s1piafqac63dyz2ycdyil4pj4ghhxx2sabm6p"))))
+    (properties
+     `((upstream-name . "org.Hs.eg.db")))
+    (build-system r-build-system)
+    (propagated-inputs
+     `(("r-annotationdbi" ,r-annotationdbi)))
+    (home-page "http://www.bioconductor.org/packages/org.Hs.eg.db/";)
+    (synopsis "Genome wide annotation for Human")
+    (description
+     "This package provides mappings from Entrez gene identifiers to various
+annotations for the human genome.")
+    (license license:artistic2.0)))
+
 (define-public r-qtl
  (package
   (name "r-qtl")
-- 
2.1.0

>From 2c56640fcefc3f2d3cd5d855fe0c716d79e1f45f Mon Sep 17 00:00:00 2001
From: Ricardo Wurmus <address@hidden>
Date: Tue, 16 Feb 2016 17:19:30 +0100
Subject: [PATCH 2/4] gnu: Add r-org-ce-eg-db.

* gnu/packages/bioinformatics.scm (r-org-ce-eg-db): New variable.
---
 gnu/packages/bioinformatics.scm | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)

diff --git a/gnu/packages/bioinformatics.scm b/gnu/packages/bioinformatics.scm
index fe82329..79f5309 100644
--- a/gnu/packages/bioinformatics.scm
+++ b/gnu/packages/bioinformatics.scm
@@ -4356,6 +4356,32 @@ genomic intervals.  In addition, it can use BAM or 
BigWig files as input.")
 annotations for the human genome.")
     (license license:artistic2.0)))
 
+(define-public r-org-ce-eg-db
+  (package
+    (name "r-org-ce-eg-db")
+    (version "3.2.3")
+    (source (origin
+              (method url-fetch)
+              ;; We cannot use bioconductor-uri here because this tarball is
+              ;; located under "data/annotation/" instead of "bioc/".
+              (uri (string-append "http://www.bioconductor.org/packages/";
+                                  "release/data/annotation/src/contrib/"
+                                  "org.Ce.eg.db_" version ".tar.gz"))
+              (sha256
+               (base32
+                "1d0lx00ybq34yqs6mziaa0lrh77xm0ggsmi76g6k95f77gi7m1sw"))))
+    (properties
+     `((upstream-name . "org.Ce.eg.db")))
+    (build-system r-build-system)
+    (propagated-inputs
+     `(("r-annotationdbi" ,r-annotationdbi)))
+    (home-page "http://www.bioconductor.org/packages/org.Ce.eg.db/";)
+    (synopsis "Genome wide annotation for Worm")
+    (description
+     "This package provides mappings from Entrez gene identifiers to various
+annotations for the C. elegans genome.")
+    (license license:artistic2.0)))
+
 (define-public r-qtl
  (package
   (name "r-qtl")
-- 
2.1.0

>From 8b7925f69c6cd69a62407990707be094e0997536 Mon Sep 17 00:00:00 2001
From: Ricardo Wurmus <address@hidden>
Date: Tue, 16 Feb 2016 17:19:55 +0100
Subject: [PATCH 3/4] gnu: Add r-org-dm-eg-db.

* gnu/packages/bioinformatics.scm (r-org-dm-eg-db): New variable.
---
 gnu/packages/bioinformatics.scm | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)

diff --git a/gnu/packages/bioinformatics.scm b/gnu/packages/bioinformatics.scm
index 79f5309..deaef00 100644
--- a/gnu/packages/bioinformatics.scm
+++ b/gnu/packages/bioinformatics.scm
@@ -4382,6 +4382,32 @@ annotations for the human genome.")
 annotations for the C. elegans genome.")
     (license license:artistic2.0)))
 
+(define-public r-org-dm-eg-db
+  (package
+    (name "r-org-dm-eg-db")
+    (version "3.2.3")
+    (source (origin
+              (method url-fetch)
+              ;; We cannot use bioconductor-uri here because this tarball is
+              ;; located under "data/annotation/" instead of "bioc/".
+              (uri (string-append "http://www.bioconductor.org/packages/";
+                                  "release/data/annotation/src/contrib/"
+                                  "org.Dm.eg.db_" version ".tar.gz"))
+              (sha256
+               (base32
+                "0mib46c7nr00l7mh290n383za9hyl91a1dc6jhjbk884jmxaxyz6"))))
+    (properties
+     `((upstream-name . "org.Dm.eg.db")))
+    (build-system r-build-system)
+    (propagated-inputs
+     `(("r-annotationdbi" ,r-annotationdbi)))
+    (home-page "http://www.bioconductor.org/packages/org.Dm.eg.db/";)
+    (synopsis "Genome wide annotation for Fly")
+    (description
+     "This package provides mappings from Entrez gene identifiers to various
+annotations for the fruit fly genome.")
+    (license license:artistic2.0)))
+
 (define-public r-qtl
  (package
   (name "r-qtl")
-- 
2.1.0

>From 3e07a3c07423754852d82c909447508d41f28c30 Mon Sep 17 00:00:00 2001
From: Ricardo Wurmus <address@hidden>
Date: Tue, 16 Feb 2016 17:20:32 +0100
Subject: [PATCH 4/4] gnu: Add r-org-mm-eg-db.

* gnu/packages/bioinformatics.scm (r-org-mm-eg-db): New variable.
---
 gnu/packages/bioinformatics.scm | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)

diff --git a/gnu/packages/bioinformatics.scm b/gnu/packages/bioinformatics.scm
index deaef00..623eeb6 100644
--- a/gnu/packages/bioinformatics.scm
+++ b/gnu/packages/bioinformatics.scm
@@ -4408,6 +4408,32 @@ annotations for the C. elegans genome.")
 annotations for the fruit fly genome.")
     (license license:artistic2.0)))
 
+(define-public r-org-mm-eg-db
+  (package
+    (name "r-org-mm-eg-db")
+    (version "3.2.3")
+    (source (origin
+              (method url-fetch)
+              ;; We cannot use bioconductor-uri here because this tarball is
+              ;; located under "data/annotation/" instead of "bioc/".
+              (uri (string-append "http://www.bioconductor.org/packages/";
+                                  "release/data/annotation/src/contrib/"
+                                  "org.Mm.eg.db_" version ".tar.gz"))
+              (sha256
+               (base32
+                "0wh1pm3npdg7070875kfgiid3bqkz3q7rq6snhk6bxfvph00298y"))))
+    (properties
+     `((upstream-name . "org.Mm.eg.db")))
+    (build-system r-build-system)
+    (propagated-inputs
+     `(("r-annotationdbi" ,r-annotationdbi)))
+    (home-page "http://www.bioconductor.org/packages/org.Mm.eg.db/";)
+    (synopsis "Genome wide annotation for Mouse")
+    (description
+     "This package provides mappings from Entrez gene identifiers to various
+annotations for the mouse genome.")
+    (license license:artistic2.0)))
+
 (define-public r-qtl
  (package
   (name "r-qtl")
-- 
2.1.0