From c6ec09ebb994e6b281a513be8465d81ed0dfc003 Mon Sep 17 00:00:00 2001
From: Efraim Flashner
Date: Thu, 10 Nov 2022 17:56:36 +0200
Subject: gn: Add pgr-tk

---
 gn/packages/bioinformatics.scm | 111 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 111 insertions(+)

(limited to 'gn/packages/bioinformatics.scm')

diff --git a/gn/packages/bioinformatics.scm b/gn/packages/bioinformatics.scm
index 2d1b55f..4e3c92b 100644
--- a/gn/packages/bioinformatics.scm
+++ b/gn/packages/bioinformatics.scm
@@ -54,6 +54,7 @@
   #:use-module (gnu packages java)
   #:use-module (gnu packages jemalloc)
   #:use-module (gnu packages linux)
+  #:use-module (gnu packages llvm)
   #:use-module (gnu packages machine-learning)
   #:use-module (gnu packages maths)
   #:use-module (gnu packages mpi)
@@ -510,6 +511,116 @@ reads.")
 collapses them into a non-redundant graph structure.")
     (license license:expat)))
 
+(define-public pgr-tk
+  (package
+    (name "pgr-tk")
+    (version "0.3.6")
+    (source
+      (origin
+        (method git-fetch)
+        (uri (git-reference
+               (url "https://github.com/Sema4-Research/pgr-tk")
+               (commit (string-append "v" version))
+               (recursive? #t)))        ; agc, WFA
+        (file-name (git-file-name name version))
+        (sha256
+         (base32 "160ngqbi8cbgaafq8crhfqv039mxr9jgl8hxxpwz0fh8mrh4003y"))
+        (snippet
+         #~(begin
+             (use-modules (guix build utils))
+             (substitute* (find-files "." "Cargo.toml")
+               ;; Only use the major+minor version to decrease the number of
+               ;; special version crates.
+               (("(.*= \")([[:digit:]]+\\.[[:digit:]]+)\\.[[:digit:]]+(\".*)"
+                 _ name version tail)
+                (string-append name version tail))
+               ;; Then fix the version string for the actual package.
+               (("^version = \".*")
+                (string-append "version = \"" #$version "\"\n")))))))
+    (build-system cargo-build-system)
+    (arguments
+     `(#:install-source? #f
+       #:cargo-test-flags
+       (list "--release" "--"
+             "--skip=get_aln_segements"
+             "--skip=get_shmmr_dots"
+             "--skip=AGCFile"
+             "--skip=SeqIndexDB")
+       #:cargo-inputs
+       (("rust-bindgen" ,rust-bindgen-0.58)
+        ("rust-bgzip" ,rust-bgzip-0.2)
+        ("rust-byteorder" ,rust-byteorder-1)
+        ("rust-clap" ,rust-clap-3.1)
+        ("rust-cuckoofilter" ,rust-cuckoofilter-0.5)
+        ("rust-flate2" ,rust-flate2-1)
+        ("rust-libc" ,rust-libc-0.2)
+        ("rust-log" ,rust-log-0.4)
+        ("rust-petgraph" ,rust-petgraph-0.6)
+        ("rust-pyo3" ,rust-pyo3-0.14)
+        ("rust-rayon" ,rust-rayon-1)
+        ("rust-regex" ,rust-regex-1)
+        ("rust-rustc-hash" ,rust-rustc-hash-1)
+        ("rust-serde" ,rust-serde-1)
+        ("rust-serde-json" ,rust-serde-json-1)
+        ("rust-simple-logger" ,rust-simple-logger-1))
+       #:phases
+       (modify-phases %standard-phases
+         (add-after 'unpack 'insert-wfa-source
+           (lambda* (#:key inputs #:allow-other-keys)
+             (copy-recursively (assoc-ref inputs "wfa-src")
+                               "rs-wfa/WFA")))
+         (add-after 'unpack 'adjust-source
+           (lambda _
+             (substitute* '("pgr-bin/build.rs"
+                            "pgr-db/build.rs"
+                            "pgr-tk/build.rs")
+               (("git") "ls")
+               (("bioconda") "Guix"))
+             ;; Build with zlib, not zlib-ng
+             (substitute* '("pgr-bin/Cargo.toml"
+                            "pgr-db/Cargo.toml")
+               (("zlib-ng-compat") "zlib"))))
+         (replace 'install
+           (lambda* (#:key outputs #:allow-other-keys)
+             (let ((out (assoc-ref outputs "out")))
+               (with-directory-excursion "target/release"
+                 (install-file "libpgrtk.so" (string-append out "/lib"))
+                 (for-each
+                   (lambda (file)
+                     (install-file file (string-append out "/bin")))
+                   (list "pgr-filter"
+                         "pgr-mdb"
+                         "pgr-multifilter"
+                         "pgr-probe-match"
+                         "pgr-shmmr-pair-count")))))))))
+    (inputs (list clang python zlib))
+    (native-inputs
+     `(("pkg-config" ,pkg-config)
+       ("wfa-src"
+        ,(origin
+           (method git-fetch)
+           (uri (git-reference
+                  ;; forPYO3 branch, 14-03-2021
+                  (url "https://github.com/cschin/WFA")
+                  (commit "1f8c8d2905ed482cd2d306a1676d60c2a45cb098")))
+           (file-name "wfa-for-pgr-tk")
+           (sha256
+            (base32 "19h1cjp2bdlcfq5c6rsbk8bc0f8zn64b471dhj4xlfxd1prv2dpk"))))))
+    (home-page "https://github.com/Sema4-Research/pgr-tk")
+    (synopsis "Pangenome Research Tool Kit")
+    (description
+     "PGR-TK provides pangenome assembly management, query and
+@acronym{Minimizer Anchored Pangenome, MAP} Graph Generation.  It is a project
+to provide Python and Rust libraries to facilitate pangenomics analysis.
+Several algorithms and data structures used for the Peregrine Genome Assembler
+are useful for Pangenomics analysis as well.  This repo takes those algorithms
+and data structure, combining other handy 3rd party tools to expose them as a
+library in Python (with Rust code for those computing parts that need
+performance.)")
+    (license (license:non-copyleft
+               "file:///LICENSE"
+               "CC-BY-NC-SA 4.0"))))
+
 (define-public pangenie
   (let ((commit "e779076827022d1416ab9fabf99a03d8f4725956") ; September 2, 2021 from phasing-tests branch
         (revision "2"))
-- 
cgit v1.2.3


From 3c387a68ac8baec058352e0933f347ac4189019d Mon Sep 17 00:00:00 2001
From: Efraim Flashner
Date: Thu, 10 Nov 2022 17:56:59 +0200
Subject: gn: Add missing imports and package licenses.

---
 gn/packages/bioinformatics.scm |  2 +-
 gn/packages/genenetwork.scm    |  1 +
 gn/packages/ruby.scm           | 54 +++++++++++++++++++++++++++---------------
 3 files changed, 37 insertions(+), 20 deletions(-)

(limited to 'gn/packages/bioinformatics.scm')

diff --git a/gn/packages/bioinformatics.scm b/gn/packages/bioinformatics.scm
index 4e3c92b..5058535 100644
--- a/gn/packages/bioinformatics.scm
+++ b/gn/packages/bioinformatics.scm
@@ -461,7 +461,7 @@ reads.")
     (description "Variant detection in massively parallel sequencing data.")
     ;; Free for non-commercial use by academic, government, and
     ;; non-profit/not-for-profit institutions
-    (license license:non-copyleft)))
+    (license (license:non-copyleft "file:///LICENSE"))))
 
 (define-public edirect-gn
   (deprecated-package "edirect-gn" edirect))
diff --git a/gn/packages/genenetwork.scm b/gn/packages/genenetwork.scm
index 4ce0d7f..e6750da 100644
--- a/gn/packages/genenetwork.scm
+++ b/gn/packages/genenetwork.scm
@@ -29,6 +29,7 @@
   #:use-module (gnu packages graphviz)
   #:use-module (gnu packages parallel)
   #:use-module (gnu packages python)
+  #:use-module (gnu packages python-build)
   #:use-module (gnu packages python-check)
   #:use-module (gnu packages python-crypto)
   #:use-module (gnu packages python-science)
diff --git a/gn/packages/ruby.scm b/gn/packages/ruby.scm
index d1df74c..79e210c 100644
--- a/gn/packages/ruby.scm
+++ b/gn/packages/ruby.scm
@@ -148,7 +148,8 @@ output")
     (description
      "Ruby/NTLM HTTP provides NTLM authentication over http.")
     (home-page "http://www.mindflowsolutions.net")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public ruby-webrobots
   (package
@@ -171,7 +172,8 @@ output")
      "This library helps write robots.txt compliant web robots in Ruby.
 ")
     (home-page "https://github.com/knu/webrobots")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public ruby-mechanize ; guix maybe ready
   (package
@@ -240,7 +242,7 @@ a history.")
 ")
   (home-page
     "https://github.com/elasticsearch/elasticsearch-ruby/tree/master/elasticsearch-transport")
-  (license #f)))
+  (license (license:non-copyleft "will fill in later"))))
 
 (define-public ruby-elasticsearch-api
 (package
@@ -266,7 +268,7 @@ a history.")
 ")
   (home-page
     "https://github.com/elasticsearch/elasticsearch-ruby/tree/master/elasticsearch-api")
-  (license #f)))
+  (license (license:non-copyleft "will fill in later"))))
 
 (define-public ruby-elasticsearch
   (package
@@ -295,7 +297,7 @@ a history.")
 ")
    (home-page
     "http://github.com/elasticsearch/elasticsearch-ruby")
-   (license #f)))
+   (license (license:non-copyleft "will fill in later"))))
 
 
 
@@ -1720,7 +1722,8 @@ and HTML without having to deal with character set issues.")
     acts as a drop-in replacement for Rails `auto_link`
 ")
     (home-page "https://github.com/vmg/rinku")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public ruby-sidekiq
   (package
@@ -1745,7 +1748,8 @@ and HTML without having to deal with character set issues.")
     (description
       "Simple, efficient background processing for Ruby.")
     (home-page "https://sidekiq.org")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public ruby-mini-scheduler
   (package
@@ -2007,7 +2011,8 @@ and HTML without having to deal with character set issues.")
     (synopsis "")
     (description "")
     (home-page "")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public generate-ninja   ; or 'gn'
   (let ((commit "dfcbc6fed0a8352696f92d67ccad54048ad182b3")
@@ -2476,7 +2481,8 @@ Use rake-compiler-dock to enter an interactive shell session or add a task to yo
     (description
       "The CSV library provides a complete interface to CSV files and data.  It offers tools to enable you to read and write to and from Strings or IO objects, as needed.")
     (home-page "https://github.com/ruby/csv")
-    (license (list #f #f))))
+    (license (list (license:non-copyleft "will fill in later")
+                   (license:non-copyleft "will fill in later")))))
 
 (define-public ruby-maxminddb
   (package
@@ -2690,7 +2696,8 @@ Use rake-compiler-dock to enter an interactive shell session or add a task to yo
     (description
       "HTTP Accept, Accept-Charset, Accept-Encoding, and Accept-Language for Ruby/Rack")
     (home-page "http://mjijackson.github.com/rack-accept")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public ruby-mustermann
  (package
@@ -2948,7 +2955,8 @@ Use rake-compiler-dock to enter an interactive shell session or add a task to yo
     (description
       "Run test/unit tests by line number.  Metal!")
     (home-page "https://github.com/qrush/m")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public ruby-allocation-stats
   (package
@@ -3027,7 +3035,8 @@ Use rake-compiler-dock to enter an interactive shell session or add a task to yo
     (description
       "FakeWeb is a helper for faking web requests in Ruby.  It works at a global level, without modifying code or writing extensive stubs.")
     (home-page "http://github.com/chrisk/fakeweb")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public ruby-simple-oauth
   (package
@@ -3364,7 +3373,8 @@ Psych also knows how to serialize and de-serialize most Ruby objects to and from
       "   Kwalify is a parser, schema validator, and data binding tool for YAML and JSON.
 ")
     (home-page "http://www.kuwata-lab.com/kwalify/")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public ruby-reek
   (package
@@ -4375,7 +4385,8 @@ non-blocking I/O methods for Ruby without raising exceptions on
 EAGAIN and EINPROGRESS.
 ")
     (home-page "https://yhbt.net/kgio/")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public ruby-unicorn
   (package
@@ -4405,7 +4416,8 @@ advantage of features in Unix/Unix-like kernels.  Slow clients should
 only be served by placing a reverse proxy capable of fully buffering
 both the the request and response in between unicorn and slow clients.")
     (home-page "https://yhbt.net/unicorn/")
-    (license (list #f #f))))
+    (license (list (license:non-copyleft "will fill in later")
+                   (license:non-copyleft "will fill in later")))))
 
 (define-public ruby-pry-rails
   (package
@@ -4463,7 +4475,8 @@ both the the request and response in between unicorn and slow clients.")
     (description
       "CSS flipper for right-to-left processing.  A Ruby port of https://github.com/ded/r2")
     (home-page "https://github.com/mzsanford/R2rb")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public ruby-openssl
   (package
@@ -5146,7 +5159,8 @@ public-key signature system described in RFC 8032.")
     (description
       "Annotates Rails/ActiveRecord Models, routes, fixtures, and others based on the database schema.")
     (home-page "https://github.com/ctran/annotate_models")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public ruby-flamegraph
   (package
@@ -5332,7 +5346,8 @@ specify.")
     (description
       "Port of arc90's readability project to ruby")
     (home-page "https://github.com/cantino/ruby-readability")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 ;; TODO: 0.3.3 never finishes the install phase with ruby-2.6
 ;; TODO: Unbundle cppjieba
@@ -5614,7 +5629,8 @@ with processing Chinese text.")
 \"test/unit/rr\".
 ")
     (home-page "https://github.com/test-unit/test-unit-rr")
-    (license #f)))
+    (license (license:non-copyleft
+               "will fill in later"))))
 
 (define-public ruby-aws-sdk-kinesis
   (package
-- 
cgit v1.2.3


From dd9aab67b68cad7f2fa2c15626d73b1e25198dee Mon Sep 17 00:00:00 2001
From: Efraim Flashner
Date: Sun, 13 Nov 2022 12:57:36 +0200
Subject: gn: pgr-tk: Unbundle agc.

---
 gn/packages/bioinformatics.scm | 119 ++++++++++++++++++++++++++++++++++++++---
 1 file changed, 113 insertions(+), 6 deletions(-)

(limited to 'gn/packages/bioinformatics.scm')

diff --git a/gn/packages/bioinformatics.scm b/gn/packages/bioinformatics.scm
index 5058535..8f98b8e 100644
--- a/gn/packages/bioinformatics.scm
+++ b/gn/packages/bioinformatics.scm
@@ -33,6 +33,7 @@
   #:use-module (gnu packages bioinformatics)
   #:use-module (gnu packages boost)
   #:use-module (gnu packages bootstrap)
+  #:use-module (gnu packages c)
   #:use-module (gnu packages check)
   #:use-module (gnu packages cmake)
   #:use-module (gnu packages compression)
@@ -511,6 +512,98 @@ reads.")
 collapses them into a non-redundant graph structure.")
     (license license:expat)))
 
+(define-public agc-for-pgr-tk
+  (let ((commit "453c0afdc54b4aa00fa8e97a63f196931fdb81c4") ; April 26, 2022
+        (revision "1"))
+    (package
+      (name "agc")
+      (version (git-version "2.0" revision commit))
+      (source
+        (origin
+          (method git-fetch)
+          (uri (git-reference
+                 (url "https://github.com/cschin/agc")
+                 (commit commit)))
+          (file-name (git-file-name name version))
+          (sha256
+           (base32 "1v5s79rl38dcyy5h1lykbp6clcbqq9winn533j54y49q1jp8chix"))
+          (snippet
+           #~(begin
+               (use-modules (guix build utils))
+               ;; Copy the two radul files we can't find a replacement for:
+               ;; https://github.com/refresh-bio/RADULS
+               (mkdir "keep-libs")
+               (rename-file "libs/raduls.h" "keep-libs/raduls.h")
+               (rename-file "libs/libraduls.a" "keep-libs/libraduls.a")
+               (delete-file-recursively "libs")
+               (rename-file "keep-libs" "libs")
+
+               (delete-file-recursively "py_agc_api/pybind11-2.8.1")
+               (substitute* '("makefile" "makefile.release")
+                 (("-mavx") "")
+                 (("-m64") "")
+                 (("\\$\\(AGC_LIBS_DIR)\\/mimalloc/\\$\\(LIB_ALLOC\\)")
+                  "$(pkg-config --cflags --libs mimalloc) /usr/lib/libmimalloc.so")
+                 (("\\$\\(AGC_LIBS_DIR)\\/\\$\\(LIB_ZLIB\\)")
+                  "$(pkg-config --cflags --libs zlib) /usr/lib/libz.so")
+                 (("\\$\\(AGC_LIBS_DIR)\\/\\$\\(LIB_ZSTD\\)")
+                  "$(pkg-config --cflags --libs libzstd) /usr/lib/libzstd.so")
+                 (("^PYBIND11_LIB = .*") "PYBIND11_LIB = /usr/include/pybind11")
+                 (("\\$\\(PYBIND11_LIB\\)/include") "$(PYBIND11_LIB)"))
+               (substitute* (find-files "src" "\\.(h|cpp)$")
+                 (("../../libs/ketopt.h") "ketopt.h")
+                 (("../../libs/zlib.h") "zlib.h")
+                 (("../../libs/zstd.h") "zstd.h"))))))
+      (build-system gnu-build-system)
+      (arguments
+       `(#:tests? #f                    ; No tests.
+         #:phases
+         (modify-phases %standard-phases
+           (delete 'configure)          ; No configure script.
+           (add-after 'unpack 'adjust-sources
+             (lambda* (#:key inputs #:allow-other-keys)
+               (let ((mimalloc (assoc-ref inputs "mimalloc")))
+                 (substitute* '("makefile" "makefile.release")
+                   (("/usr/include/pybind11")
+                    (search-input-directory inputs "/include/pybind11"))
+                   (("/usr/lib/libmimalloc.so")
+                    (search-input-file inputs "/lib/libmimalloc.so"))
+                   (("/usr/lib/libz.so")
+                    (search-input-file inputs "/lib/libz.so"))
+                   (("/usr/lib/libzstd.so")
+                    (search-input-file inputs "/lib/libzstd.so"))
+                   (("pkg-config") ,(pkg-config-for-target))))))
+           (replace 'install
+             (lambda* (#:key outputs #:allow-other-keys)
+               (let* ((out (assoc-ref outputs "out"))
+                      (include (string-append out "/include/")))
+                 (install-file "agc" (string-append out "/bin"))
+                 (install-file "libagc.so" (string-append out "/lib"))
+                 (mkdir-p (string-append include "app"))
+                 (mkdir-p (string-append include "core"))
+                 (mkdir-p (string-append include "lib-cxx"))
+                 (with-directory-excursion "src"
+                   (for-each
+                     (lambda (file)
+                       (copy-file file (string-append include file)))
+                     (find-files "." "\\.h$")))))))))
+      (native-inputs
+       (list minimap2                   ; for ketopt.h
+             pkg-config))
+      (inputs
+       (list mimalloc
+             python
+             pybind11
+             zlib
+             (list zstd "lib")))
+      (home-page "https://github.com/cschin/agc")
+      (synopsis "Assembled Genomes Compressor")
+      (description
+       "@acronym{Assembled Genomes Compressor, AGC} is a tool designed to
+compress collections of de-novo assembled genomes.  It can be used for various
+types of datasets: short genomes (viruses) as well as long (humans).")
+      (license license:expat))))
+
 (define-public pgr-tk
   (package
     (name "pgr-tk")
@@ -520,11 +613,10 @@ collapses them into a non-redundant graph structure.")
         (method git-fetch)
         (uri (git-reference
                (url "https://github.com/Sema4-Research/pgr-tk")
-               (commit (string-append "v" version))
-               (recursive? #t)))        ; agc, WFA
+               (commit (string-append "v" version))))
         (file-name (git-file-name name version))
         (sha256
-         (base32 "160ngqbi8cbgaafq8crhfqv039mxr9jgl8hxxpwz0fh8mrh4003y"))
+         (base32 "0vm1k63v91zd0pfbg2zmwskajylz8xg83m63qxwaiwny5f4y6f1j"))
         (snippet
          #~(begin
              (use-modules (guix build utils))
@@ -570,7 +662,7 @@ collapses them into a non-redundant graph structure.")
              (copy-recursively (assoc-ref inputs "wfa-src")
                                "rs-wfa/WFA")))
          (add-after 'unpack 'adjust-source
-           (lambda _
+           (lambda* (#:key inputs #:allow-other-keys)
              (substitute* '("pgr-bin/build.rs"
                             "pgr-db/build.rs"
                             "pgr-tk/build.rs")
@@ -579,7 +671,17 @@ collapses them into a non-redundant graph structure.")
              ;; Build with zlib, not zlib-ng
              (substitute* '("pgr-bin/Cargo.toml"
                             "pgr-db/Cargo.toml")
-               (("zlib-ng-compat") "zlib"))))
+               (("zlib-ng-compat") "zlib"))
+             ;; Don't look for agc to be bundled.
+             (substitute* "pgr-db/wrapper.h"
+               (("../agc/src/lib-cxx/agc-api.h") "lib-cxx/agc-api.h"))
+             (substitute* "pgr-db/build.rs"
+               ((".*panic!\\(\"Error.*") ""))
+             (mkdir-p "target/release")
+             (symlink (search-input-file inputs "/bin/agc")
+                      "target/release/agc")
+             (symlink (search-input-file inputs "/lib/libagc.so")
+                      "target/release/libagc")))
          (replace 'install
            (lambda* (#:key outputs #:allow-other-keys)
              (let ((out (assoc-ref outputs "out")))
@@ -593,7 +695,12 @@ collapses them into a non-redundant graph structure.")
                          "pgr-multifilter"
                          "pgr-probe-match"
                          "pgr-shmmr-pair-count")))))))))
-    (inputs (list clang python zlib))
+    (inputs
+     (list agc-for-pgr-tk
+           clang
+           python
+           zlib
+           (list zstd "lib")))
     (native-inputs
      `(("pkg-config" ,pkg-config)
        ("wfa-src"
-- 
cgit v1.2.3


From cc8ee17a6e63cf20adc9374735a1322dd9c7f88e Mon Sep 17 00:00:00 2001
From: Efraim Flashner
Date: Wed, 16 Nov 2022 14:56:36 +0200
Subject: gn: Add graph-genotyper and dependent packages

---
 gn/packages/bioinformatics.scm | 133 ++++++++++++++++++++++++++++++++++++++++
 gn/packages/crates-io.scm      | 135 ++++++++++++++++++++++++++++++++++++++++-
 2 files changed, 267 insertions(+), 1 deletion(-)

(limited to 'gn/packages/bioinformatics.scm')

diff --git a/gn/packages/bioinformatics.scm b/gn/packages/bioinformatics.scm
index 8f98b8e..76b93d9 100644
--- a/gn/packages/bioinformatics.scm
+++ b/gn/packages/bioinformatics.scm
@@ -512,6 +512,34 @@ reads.")
 collapses them into a non-redundant graph structure.")
     (license license:expat)))
 
+(define-public gafpack
+  (let ((commit "ad31875b6914d964c6fd72d1bf334f0843538fb6")     ; November 10, 2022
+        (revision "1"))
+    (package
+      (name "gafpack")
+      (version (git-version "0.0.0" revision commit))
+      (source
+        (origin
+          (method git-fetch)
+          (uri (git-reference
+                 (url "https://github.com/ekg/gafpack")
+                 (commit commit)))
+          (file-name (git-file-name name version))
+          (sha256
+           (base32 "0di2psh0ls7jlbnqs7k71p55f73pn23a09k1h3ril7gwjcrzr3rk"))))
+      (build-system cargo-build-system)
+      (arguments
+       `(#:install-source? #f
+         #:cargo-inputs
+         (("rust-clap" ,rust-clap-4)
+          ("rust-gfa" ,rust-gfa-0.10))))
+      (home-page "https://github.com/ekg/gafpack")
+      (synopsis "Convert variation graph alignments to coverage maps over nodes")
+      (description
+       "Gafpack converts alignments to pangenome variation graphs to coverage
+maps useful in haplotype-based genotyping.")
+      (license license:expat))))
+
 (define-public agc-for-pgr-tk
   (let ((commit "453c0afdc54b4aa00fa8e97a63f196931fdb81c4") ; April 26, 2022
         (revision "1"))
@@ -728,6 +756,61 @@ performance.)")
                "file:///LICENSE"
                "CC-BY-NC-SA 4.0"))))
 
+(define-public graph-genotyper
+  (let ((commit "e7cc6b43a5b1f389d76bf9aac7f2ee02f92caeaf") ; October 17, 2022
+        (revision "13"))
+    (package
+      (name "graph-genotyper")
+      (version (git-version "0.0.0" revision commit))
+      (source (origin
+        (method git-fetch)
+        (uri (git-reference
+               (url "https://github.com/davidebolo1993/graph_genotyper")
+               (commit commit)))
+        (file-name (git-file-name name version))
+        (sha256
+         (base32 "1l8yjpkqamiqr1q5i7vr5z04aba7skpbcwyc9dx5fiklvljjfhcx"))))
+      (build-system copy-build-system)
+      (arguments
+       `(#:install-plan
+         '(("genotype.py" "bin/")
+           ("genotype.sh" "bin/"))
+         #:phases
+         (modify-phases %standard-phases
+           (add-after 'install 'wrap-genotype
+             (lambda* (#:key inputs outputs #:allow-other-keys)
+               (let ((out (assoc-ref outputs "out")))
+                 (wrap-script (string-append out "/bin/genotype.sh")
+                  `("GUIX_PYTHONPATH" ":" prefix (,(getenv "GUIX_PYTHONPATH")))
+                  `("PATH" ":" prefix
+                    ,(map (lambda (file-name)
+                            (string-append (assoc-ref inputs file-name) "/bin"))
+                          (list "gafpack"
+                                "odgi"
+                                "python"
+                                "samtools"
+                                "vg"))))))))))
+      (inputs
+       (list gafpack
+             guile-3.0
+             odgi
+             python
+             python-numpy
+             python-pandas
+             python-scipy
+             samtools
+             vg))
+      (home-page "https://bitbucket.org/jana_ebler")
+      (synopsis "Genotyping based on k-mers and pangenome graphs")
+      (description
+       "This package provides a genotyper for various types of genetic variants
+(such as SNPs, indels and structural variants).  Genotypes are computed based on
+read k-mer counts and a panel of known haplotypes.  A description of the method
+can be found @url{https://www.biorxiv.org/content/10.1101/2020.11.11.378133v1,
+here}.")
+      (license (license:non-copyleft
+                 "No license listed")))))
+
 (define-public pangenie
   (let ((commit "e779076827022d1416ab9fabf99a03d8f4725956") ; September 2, 2021 from phasing-tests branch
         (revision "2"))
@@ -1811,6 +1894,56 @@ available to other researchers.")
                  (delete-file "scanpy/tests/test_pca.py")
                  #t)))))))))
 
+;; TODO: Unbundle everything
+(define-public odgi
+  (package
+    (name "odgi")
+    (version "0.8.1")
+    (source (origin
+              (method url-fetch)
+              (uri (string-append "https://github.com/pangenome/odgi/releases"
+                                  "/download/v" version
+                                  "/odgi-v" version ".tar.gz"))
+              (sha256
+               (base32 "175083pb9hp0vn9a00hbxlayyk5a5j8p52yq5qfmbnfvndisbmbv"))
+              (snippet
+               #~(begin
+                   (use-modules (guix build utils))
+                   (substitute* "CMakeLists.txt"
+                     (("-march=native") "")
+                     (("-msse4\\.2") ""))
+                   (delete-file-recursively "deps/pybind11")
+                   (delete-file-recursively "deps/sdsl-lite")))))
+    (build-system cmake-build-system)
+    (native-inputs
+     (list pkg-config))
+    (inputs
+     (list jemalloc
+           libdivsufsort
+           pybind11
+           python
+           sdsl-lite))
+    (home-page "https://github.com/vgteam/odgi")
+    (synopsis "Optimized Dynamic Genome/Graph Implementation")
+    (description "@acronym{Optimized Dynamic Genome/Graph Implementation, odgi}
+provides an efficient and succinct dynamic DNA sequence graph model, as well as
+a host of algorithms that allow the use of such graphs in bioinformatic
+analyses.
+
+Careful encoding of graph entities allows odgi to efficiently compute and
+transform pangenomes with minimal overheads.  @command{odgi} implements a
+dynamic data structure that leveraged multi-core CPUs and can be updated on the
+fly.
+
+The edges and path steps are recorded as deltas between the current node id and
+the target node id, where the node id corresponds to the rank in the global
+array of nodes.  Graphs built from biological data sets tend to have local
+partial order and, when sorted, the deltas be small.  This allows them to be
+compressed with a variable length integer representation, resulting in a small
+in-memory footprint at the cost of packing and unpacking.")
+    (properties '((tunable? . #t)))
+    (license license:expat)))
+
 (define-public vg
   (package
     (name "vg")
diff --git a/gn/packages/crates-io.scm b/gn/packages/crates-io.scm
index 847b2ef..8b7bc4c 100644
--- a/gn/packages/crates-io.scm
+++ b/gn/packages/crates-io.scm
@@ -7,7 +7,8 @@
   #:use-module (guix build-system cargo)
   #:use-module (gnu packages crates-graphics)
   #:use-module (gnu packages crates-io)
-  #:use-module (gnu packages maths))
+  #:use-module (gnu packages maths)
+  #:use-module (gnu packages python))
 
 (define-public rust-bgzip-0.2
   (package
@@ -65,6 +66,25 @@ for massive key sets}.  It generates an @acronym{MPHF, minimal perfect hash
 functions} for a collection of hashable objects.")
     (license license:expat)))
 
+(define-public rust-clap-lex-0.3
+  (package
+    (name "rust-clap-lex")
+    (version "0.3.0")
+    (source (origin
+              (method url-fetch)
+              (uri (crate-uri "clap-lex" version))
+              (file-name (string-append name "-" version ".tar.gz"))
+              (sha256
+               (base32
+                "1a4dzbnlxiamfsn0pnkhn7n9bdfjh66j9fxm6mmr7d227vvrhh8d"))))
+    (build-system cargo-build-system)
+    (arguments
+     `(#:cargo-inputs (("rust-os-str-bytes" ,rust-os-str-bytes-6))))
+    (home-page "https://github.com/clap-rs/clap/tree/master/clap_lex")
+    (synopsis "Minimal, flexible command line parser")
+    (description "Minimal, flexible command line parser")
+    (license (list license:expat license:asl2.0))))
+
 (define-public rust-cuckoofilter-0.5
   (package
     (name "rust-cuckoofilter")
@@ -352,6 +372,51 @@ Python code from a Rust binary is also supported.")
     (description "This package provides succinct data structures for Rust.")
     (license (list license:expat license:asl2.0))))
 
+(define-public rust-terminal-size-0.2
+  (package
+    (name "rust-terminal-size")
+    (version "0.2.2")
+    (source (origin
+              (method url-fetch)
+              (uri (crate-uri "terminal-size" version))
+              (file-name (string-append name "-" version ".tar.gz"))
+              (sha256
+               (base32
+                "0yhza8sc6jkka6j0nq5sl749ckx1jagvxp3b38yhh4px6k291jj0"))))
+    (build-system cargo-build-system)
+    (arguments
+     `(#:skip-build? #t
+       ;#:cargo-inputs
+       ;(("rust-rustix" ,rust-rustix-0.35)
+       ; ("rust-windows-sys" ,rust-windows-sys-0.42))
+       ))
+    (home-page "https://github.com/eminence/terminal-size")
+    (synopsis "Gets the size of your Linux or Windows terminal")
+    (description "Gets the size of your Linux or Windows terminal")
+    (license (list license:expat license:asl2.0))))
+
+(define-public rust-unic-emoji-char-0.9
+  (package
+    (name "rust-unic-emoji-char")
+    (version "0.9.0")
+    (source (origin
+              (method url-fetch)
+              (uri (crate-uri "unic-emoji-char" version))
+              (file-name (string-append name "-" version ".tar.gz"))
+              (sha256
+               (base32
+                "0ka9fr7s6lv0z43r9xphg9injn35pfxf9g9q18ki0wl9d0g241qb"))))
+    (build-system cargo-build-system)
+    (arguments
+     `(#:cargo-inputs
+       (("rust-unic-char-property" ,rust-unic-char-property-0.9)
+        ("rust-unic-char-range" ,rust-unic-char-range-0.9)
+        ("rust-unic-ucd-version" ,rust-unic-ucd-version-0.9))))
+    (home-page "https://github.com/open-i18n/rust-unic/")
+    (synopsis "UNIC â Unicode Emoji â Emoji Character Properties")
+    (description "UNIC â Unicode Emoji â Emoji Character Properties")
+    (license (list license:expat license:asl2.0))))
+
 (define-public rust-clap-for-jrep
   (package
     (name "rust-clap")
@@ -499,6 +564,48 @@ or any combination.")
       Argument Parser")
     (license (list license:expat license:asl2.0))))
 
+(define-public rust-clap-4
+  (package
+    (name "rust-clap")
+    (version "4.0.9")
+    (source (origin
+              (method url-fetch)
+              (uri (crate-uri "clap" version))
+              (file-name (string-append name "-" version ".tar.gz"))
+              (sha256
+               (base32
+                "1w0dxqzrh449s9l2k8g66pdsff02599bwi5mh0gny3227kcpsq1h"))))
+    (build-system cargo-build-system)
+    (arguments
+     `(#:cargo-inputs
+       (("rust-atty" ,rust-atty-0.2)
+        ("rust-backtrace" ,rust-backtrace-0.3)
+        ("rust-bitflags" ,rust-bitflags-1)
+        ("rust-clap-derive" ,rust-clap-derive-4)
+        ("rust-clap-lex" ,rust-clap-lex-0.3)
+        ("rust-once-cell" ,rust-once-cell-1)
+        ("rust-strsim" ,rust-strsim-0.10)
+        ("rust-termcolor" ,rust-termcolor-1)
+        ("rust-terminal-size" ,rust-terminal-size-0.2)
+        ("rust-unicase" ,rust-unicase-2)
+        ("rust-unicode-width" ,rust-unicode-width-0.1))
+       #:cargo-development-inputs
+       (("rust-humantime" ,rust-humantime-2)
+        ("rust-rustversion" ,rust-rustversion-1)
+        ("rust-shlex" ,rust-shlex-1)
+        ("rust-snapbox" ,rust-snapbox-0.4)
+        ("rust-static-assertions" ,rust-static-assertions-1)
+        ("rust-trybuild" ,rust-trybuild-1)
+        ("rust-trycmd" ,rust-trycmd-0.13)
+        ("rust-unic-emoji-char" ,rust-unic-emoji-char-0.9))))
+    (home-page "https://github.com/clap-rs/clap")
+    (synopsis
+      "A simple to use, efficient, and full-featured Command Line Argument Parser")
+    (description
+      "This package provides a simple to use, efficient, and full-featured Command Line
+      Argument Parser")
+    (license (list license:expat license:asl2.0))))
+
 ;; ready to upstream, WITH rust-clap-derive
 ;; replace fields with those from upstream.
 (define-public rust-clap-derive-3.1
@@ -527,6 +634,32 @@ or any combination.")
       "Parse command line argument by defining a struct, derive crate.")
     (license (list license:expat license:asl2.0))))
 
+(define-public rust-clap-derive-4
+  (package
+    (name "rust-clap-derive")
+    (version "4.0.9")
+    (source (origin
+              (method url-fetch)
+              (uri (crate-uri "clap-derive" version))
+              (file-name (string-append name "-" version ".tar.gz"))
+              (sha256
+               (base32
+                "02zhbbmyz3dpy9ml6xfp7i8p3ffj1djvkdnkg6gr6d0s5r4hg8x4"))))
+    (build-system cargo-build-system)
+    (arguments
+     `(#:cargo-inputs
+       (("rust-heck" ,rust-heck-0.4)
+        ("rust-proc-macro-error" ,rust-proc-macro-error-1)
+        ("rust-proc-macro2" ,rust-proc-macro2-1)
+        ("rust-quote" ,rust-quote-1)
+        ("rust-syn" ,rust-syn-1))))
+    (home-page "https://github.com/clap-rs/clap/tree/master/clap_derive")
+    (synopsis
+      "Parse command line argument by defining a struct, derive crate.")
+    (description
+      "Parse command line argument by defining a struct, derive crate.")
+    (license (list license:expat license:asl2.0))))
+
 (define-public rust-textwrap-0.15
   (package
     (name "rust-textwrap")
-- 
cgit v1.2.3


From a8481f8fda599be8af4c14663cc57e538f8117c3 Mon Sep 17 00:00:00 2001
From: Efraim Flashner
Date: Wed, 30 Nov 2022 10:54:03 +0200
Subject: gn: Remove upstreamed packages

---
 gn/packages/bioinformatics.scm |   4 +-
 gn/packages/crates-io.scm      | 151 -----------------------------------------
 2 files changed, 2 insertions(+), 153 deletions(-)

(limited to 'gn/packages/bioinformatics.scm')

diff --git a/gn/packages/bioinformatics.scm b/gn/packages/bioinformatics.scm
index 76b93d9..19cc1ff 100644
--- a/gn/packages/bioinformatics.scm
+++ b/gn/packages/bioinformatics.scm
@@ -489,7 +489,7 @@ reads.")
     (arguments
      `(#:install-source? #f
        #:cargo-inputs
-       (("rust-clap" ,rust-clap-3.1)
+       (("rust-clap" ,rust-clap-3)
         ("rust-rustc-hash" ,rust-rustc-hash-1)
         ("rust-regex" ,rust-regex-1)
         ("rust-handlegraph" ,rust-handlegraph-0.7)
@@ -670,7 +670,7 @@ types of datasets: short genomes (viruses) as well as long (humans).")
        (("rust-bindgen" ,rust-bindgen-0.58)
         ("rust-bgzip" ,rust-bgzip-0.2)
         ("rust-byteorder" ,rust-byteorder-1)
-        ("rust-clap" ,rust-clap-3.1)
+        ("rust-clap" ,rust-clap-3)
         ("rust-cuckoofilter" ,rust-cuckoofilter-0.5)
         ("rust-flate2" ,rust-flate2-1)
         ("rust-libc" ,rust-libc-0.2)
diff --git a/gn/packages/crates-io.scm b/gn/packages/crates-io.scm
index 60b6229..342aca4 100644
--- a/gn/packages/crates-io.scm
+++ b/gn/packages/crates-io.scm
@@ -372,51 +372,6 @@ Python code from a Rust binary is also supported.")
     (description "This package provides succinct data structures for Rust.")
     (license (list license:expat license:asl2.0))))
 
-(define-public rust-terminal-size-0.2
-  (package
-    (name "rust-terminal-size")
-    (version "0.2.2")
-    (source (origin
-              (method url-fetch)
-              (uri (crate-uri "terminal-size" version))
-              (file-name (string-append name "-" version ".tar.gz"))
-              (sha256
-               (base32
-                "0yhza8sc6jkka6j0nq5sl749ckx1jagvxp3b38yhh4px6k291jj0"))))
-    (build-system cargo-build-system)
-    (arguments
-     `(#:skip-build? #t
-       ;#:cargo-inputs
-       ;(("rust-rustix" ,rust-rustix-0.35)
-       ; ("rust-windows-sys" ,rust-windows-sys-0.42))
-       ))
-    (home-page "https://github.com/eminence/terminal-size")
-    (synopsis "Gets the size of your Linux or Windows terminal")
-    (description "Gets the size of your Linux or Windows terminal")
-    (license (list license:expat license:asl2.0))))
-
-(define-public rust-unic-emoji-char-0.9
-  (package
-    (name "rust-unic-emoji-char")
-    (version "0.9.0")
-    (source (origin
-              (method url-fetch)
-              (uri (crate-uri "unic-emoji-char" version))
-              (file-name (string-append name "-" version ".tar.gz"))
-              (sha256
-               (base32
-                "0ka9fr7s6lv0z43r9xphg9injn35pfxf9g9q18ki0wl9d0g241qb"))))
-    (build-system cargo-build-system)
-    (arguments
-     `(#:cargo-inputs
-       (("rust-unic-char-property" ,rust-unic-char-property-0.9)
-        ("rust-unic-char-range" ,rust-unic-char-range-0.9)
-        ("rust-unic-ucd-version" ,rust-unic-ucd-version-0.9))))
-    (home-page "https://github.com/open-i18n/rust-unic/")
-    (synopsis "UNIC â Unicode Emoji â Emoji Character Properties")
-    (description "UNIC â Unicode Emoji â Emoji Character Properties")
-    (license (list license:expat license:asl2.0))))
-
 (define-public rust-clap-for-jrep
   (package
     (name "rust-clap")
@@ -520,50 +475,6 @@ or any combination.")
       (description "Rust CLI tools for manipulation of Jupyter Notebooks.")
       (license #f)))) ; There is no license.
 
-;; replace fields with those from upstream
-(define-public rust-clap-3.1
-  (package
-    (name "rust-clap")
-    (version "3.1.6")
-    (source
-      (origin
-        (method url-fetch)
-        (uri (crate-uri "clap" version))
-        (file-name (string-append name "-" version ".tar.gz"))
-        (sha256
-         (base32 "08q1hkksfixybnrwrpm44xq028wbn9yr2hnzrax9hihyq8v39jfq"))))
-    (build-system cargo-build-system)
-    (arguments
-     `(#:cargo-inputs
-       (("rust-atty" ,rust-atty-0.2)
-        ("rust-backtrace" ,rust-backtrace-0.3)
-        ("rust-bitflags" ,rust-bitflags-1)
-        ("rust-clap-derive" ,rust-clap-derive-3.1)
-        ("rust-indexmap" ,rust-indexmap-1)
-        ("rust-lazy-static" ,rust-lazy-static-1)
-        ("rust-os-str-bytes" ,rust-os-str-bytes-6)
-        ("rust-regex" ,rust-regex-1)
-        ("rust-strsim" ,rust-strsim-0.10)
-        ("rust-termcolor" ,rust-termcolor-1)
-        ("rust-terminal-size" ,rust-terminal-size-0.1)
-        ("rust-textwrap" ,rust-textwrap-0.15)
-        ("rust-unicase" ,rust-unicase-2)
-        ("rust-yaml-rust" ,rust-yaml-rust-0.4))
-       #:cargo-development-inputs
-       (("rust-criterion" ,rust-criterion-0.3)
-        ("rust-lazy-static" ,rust-lazy-static-1)
-        ("rust-regex" ,rust-regex-1)
-        ("rust-rustversion" ,rust-rustversion-1)
-        ("rust-trybuild" ,rust-trybuild-1)
-        ("rust-trycmd" ,rust-trycmd-0.12))))
-    (home-page "https://github.com/clap-rs/clap")
-    (synopsis
-      "A simple to use, efficient, and full-featured Command Line Argument Parser")
-    (description
-      "This package provides a simple to use, efficient, and full-featured Command Line
-      Argument Parser")
-    (license (list license:expat license:asl2.0))))
-
 (define-public rust-clap-4
   (package
     (name "rust-clap")
@@ -608,34 +519,6 @@ or any combination.")
       Argument Parser")
     (license (list license:expat license:asl2.0))))
 
-;; ready to upstream, WITH rust-clap-derive
-;; replace fields with those from upstream.
-(define-public rust-clap-derive-3.1
-  (package
-    (name "rust-clap-derive")
-    (version "3.1.4")
-    (source
-      (origin
-        (method url-fetch)
-        (uri (crate-uri "clap-derive" version))
-        (file-name (string-append name "-" version ".tar.gz"))
-        (sha256
-         (base32 "05mz2y6k73wc1gvv9r4mllfqslzvlwkvx77lk7769ag1xlwd15fs"))))
-    (build-system cargo-build-system)
-    (arguments
-     `(#:cargo-inputs
-       (("rust-heck" ,rust-heck-0.4)
-        ("rust-proc-macro-error" ,rust-proc-macro-error-1)
-        ("rust-proc-macro2" ,rust-proc-macro2-1)
-        ("rust-quote" ,rust-quote-1)
-        ("rust-syn" ,rust-syn-1))))
-    (home-page "https://github.com/clap-rs/clap/tree/master/clap_derive")
-    (synopsis
-      "Parse command line argument by defining a struct, derive crate.")
-    (description
-      "Parse command line argument by defining a struct, derive crate.")
-    (license (list license:expat license:asl2.0))))
-
 (define-public rust-clap-derive-4
   (package
     (name "rust-clap-derive")
@@ -662,40 +545,6 @@ or any combination.")
       "Parse command line argument by defining a struct, derive crate.")
     (license (list license:expat license:asl2.0))))
 
-(define-public rust-textwrap-0.15
-  (package
-    (name "rust-textwrap")
-    (version "0.15.0")
-    (source
-      (origin
-        (method url-fetch)
-        (uri (crate-uri "textwrap" version))
-        (file-name (string-append name "-" version ".tar.gz"))
-        (sha256
-         (base32 "1yw513k61lfiwgqrfvsjw1a5wpvm0azhpjr2kr0jhnq9c56is55i"))))
-    (build-system cargo-build-system)
-    (arguments
-     `(#:skip-build? #t     ; Not all inputs packaged
-       ;#:tests? #f          ; Skip tests for now
-       #:cargo-inputs
-       (("rust-hyphenation" ,rust-hyphenation-0.8)
-        ("rust-smawk" ,rust-smawk-0.3)
-        ("rust-terminal-size" ,rust-terminal-size-0.1)
-        ("rust-unicode-linebreak" ,rust-unicode-linebreak-0.1)
-        ("rust-unicode-width" ,rust-unicode-width-0.1))
-       #:cargo-development-inputs
-       (("rust-criterion" ,rust-criterion-0.3)
-        ;("rust-lipsum" ,rust-lipsum-0.8)
-        ("rust-termion" ,rust-termion-1)
-        ;("rust-unic-emoji-char" ,rust-unic-emoji-char-0.9)
-        ("rust-version-sync" ,rust-version-sync-0.9))))
-    (home-page "https://github.com/mgeisler/textwrap")
-    (synopsis
-      "Powerful library for word wrapping, indenting, and dedenting strings")
-    (description
-      "Powerful library for word wrapping, indenting, and dedenting strings")
-    (license license:expat)))
-
 (define-public rust-trycmd-0.12
   (package
     (name "rust-trycmd")
-- 
cgit v1.2.3


From 9fb27d3e8cd7f4ce1040bbf815b5c9031dbce3b0 Mon Sep 17 00:00:00 2001
From: Pjotr Prins
Date: Thu, 8 Dec 2022 22:24:28 +0000
Subject: Minor fixes aiming for genenetwork1

---
 gn/packages/bioinformatics.scm | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'gn/packages/bioinformatics.scm')

diff --git a/gn/packages/bioinformatics.scm b/gn/packages/bioinformatics.scm
index 19cc1ff..3cd1e1c 100644
--- a/gn/packages/bioinformatics.scm
+++ b/gn/packages/bioinformatics.scm
@@ -2406,7 +2406,7 @@ The Genome Browser itself does not draw conclusions; rather, it collates all
 relevant information in one location, leaving the exploration and interpretation
 to the user.")
     (license (list
-               license:bsd-0    ; kent/src/{utils,lib,inc,tabStorm,parasol,hg/ausoSql,hg/autoXml}
+               ;; license:bsd-0    ; kent/src/{utils,lib,inc,tabStorm,parasol,hg/ausoSql,hg/autoXml}
                license:bsd-3    ; these two for bundled htslib-1.3
                license:expat
                (license:non-copyleft
-- 
cgit v1.2.3


From 0bf0d2684677bc40c0759cb1640659bf03f882ae Mon Sep 17 00:00:00 2001
From: Efraim Flashner
Date: Wed, 22 Mar 2023 13:24:10 +0200
Subject: Add python-pytest-runner-4.

* gn/packages/python.scm (python-pytest-runner-4): New variable.
* gn/packages/bioinformatics.scm (python-pytest-runner-2): MOve to
python.scm
---
 gn/packages/bioinformatics.scm | 11 -----------
 gn/packages/python.scm         | 23 +++++++++++++++++++++++
 2 files changed, 23 insertions(+), 11 deletions(-)

(limited to 'gn/packages/bioinformatics.scm')

diff --git a/gn/packages/bioinformatics.scm b/gn/packages/bioinformatics.scm
index 3cd1e1c..b58d97e 100644
--- a/gn/packages/bioinformatics.scm
+++ b/gn/packages/bioinformatics.scm
@@ -1791,17 +1791,6 @@ reads, also called read-based phasing or haplotype assembly.  It is especially
 suitable for long reads, but works also well with short reads.")
     (license license:expat)))
 
-(define-public python-pytest-runner-2
-  (package
-    (inherit python-pytest-runner)
-    (version "2.12.2")
-    (source (origin
-              (method url-fetch)
-              (uri (pypi-uri "pytest-runner" version))
-              (sha256
-               (base32
-                "11ivjj9hfphkv4yfb2g74av4yy86y8gcbf7gbif0p1hcdfnxg3w6"))))))
-
 (define-public bh20-seq-resource
   (let ((commit "ae4cb3c2cf7103bbc84f52618bb755d7ce25775b")
         (revision "3"))
diff --git a/gn/packages/python.scm b/gn/packages/python.scm
index f3fa50b..169fac7 100644
--- a/gn/packages/python.scm
+++ b/gn/packages/python.scm
@@ -1324,6 +1324,29 @@ sorted order.")
        ("python-requests" ,python-requests)
        ,@(package-native-inputs python-pytest)))))
 
+(define-public python-pytest-runner-4
+  (package
+    (inherit python-pytest-runner)
+    (name "python-pytest-runner")
+    (version "4.5.1")
+    (source (origin
+              (method url-fetch)
+              (uri (pypi-uri "pytest-runner" version))
+              (sha256
+               (base32
+                "1vzilbayx5mznsdm1r258m3616374p6kvhsbj4j6238j9djkvjyi"))))))
+
+(define-public python-pytest-runner-2
+  (package
+    (inherit python-pytest-runner)
+    (version "2.12.2")
+    (source (origin
+              (method url-fetch)
+              (uri (pypi-uri "pytest-runner" version))
+              (sha256
+               (base32
+                "11ivjj9hfphkv4yfb2g74av4yy86y8gcbf7gbif0p1hcdfnxg3w6"))))))
+
 (define-public python-pandas-plink
   (package
     (name "python-pandas-plink")
-- 
cgit v1.2.3


From 3cddd10e7b3a52d1d3f8411417e3b59a1ae558c9 Mon Sep 17 00:00:00 2001
From: Efraim Flashner
Date: Wed, 22 Mar 2023 13:25:19 +0200
Subject: bh20-seq-resource: Update to 1.0-4.2ae7191.

* gn/packages/bioinformatics.scm (bh20-seq-resource): Update to
1.0-4.2ae7191.
[source]: Adjust snippet for changes in source.
[arguments]: Skip tests.
[propagated-inputs]: Add python-shexc-0.7.  Remove python-dateutil,
python-flask, python-pycurl, python38-ruaml.yaml-0.15.76, clustalw,
python-twint.
[native-inputs]: Remove git, python-oauth2client, python-uritemplate.
Replace python-pytest-4 with python-pytest-5, python-pytest-runner-2
with python-pytest-runner-4.
---
 gn/packages/bioinformatics.scm | 50 ++++++++++++++++++++----------------------
 1 file changed, 24 insertions(+), 26 deletions(-)

(limited to 'gn/packages/bioinformatics.scm')

diff --git a/gn/packages/bioinformatics.scm b/gn/packages/bioinformatics.scm
index b58d97e..48f1c40 100644
--- a/gn/packages/bioinformatics.scm
+++ b/gn/packages/bioinformatics.scm
@@ -1792,8 +1792,8 @@ suitable for long reads, but works also well with short reads.")
     (license license:expat)))
 
 (define-public bh20-seq-resource
-  (let ((commit "ae4cb3c2cf7103bbc84f52618bb755d7ce25775b")
-        (revision "3"))
+  (let ((commit "2ae71911cd87ce4f2eabdff21e538267b3270d45")
+        (revision "4"))
     (package
       (name "bh20-seq-resource")
       (version (git-version "1.0" revision commit))
@@ -1804,36 +1804,34 @@ suitable for long reads, but works also well with short reads.")
                        (commit commit)))
                 (file-name (git-file-name name version))
                 (sha256
-                 (base32 "1k0gsz4yc8l5znanzd094g2jp40ksbpa9667zr31ayrjx6labz02"))
+                 (base32 "1k6cc88hrcm77jwpdk2084q0zirv2vlbz3c07nmpbhk1lhqk5x0n"))
                 (modules '((guix build utils)))
                 (snippet
                  '(begin
-                    (substitute* "setup.py"
-                      (("py-dateutil") "python-dateutil"))
-                    #t))))
+                    (delete-file "gittaggers.py")))))
       (build-system python-build-system)
+      (arguments
+       (list
+         #:tests? #f))  ; Tests can't find pytest
       (propagated-inputs
-       `(("python-arvados-python-client" ,python-arvados-python-client)
-         ("python-dateutil" ,python-dateutil)
-         ("python-flask" ,python-flask)
-         ("python-magic" ,python-magic)
-         ("python-pyyaml" ,python-pyyaml)
-         ("python-pycurl" ,python-pycurl)
-         ("python-pyshex" ,python-pyshex)
-         ("python-redis" ,python-redis)
-         ("python-ruaml.yaml" ,python38-ruaml.yaml-0.15.76)
-         ("clustalw" ,clustalw)
-         ("python-schema-salad" ,python-schema-salad)
-         ("python-twint" ,python-twint)
-         ;; and for the service
-         ("python" ,python)
-         ("gunicorn" ,gunicorn)))
+       (list python-arvados-python-client
+             python-schema-salad
+             python-magic
+             python-pyshex
+             python-pyshexc-0.7
+             python-py-dateutil
+
+             ;; for the web
+             python-flask
+             python-pyyaml
+             python-redis
+
+             ;; and for the service
+             python
+             gunicorn))
       (native-inputs
-       `(("git" ,(@ (gnu packages version-control) git))
-         ("python-oauth2client" ,python-oauth2client)
-         ("python-pytest" ,python-pytest-4)
-         ("python-pytest-runner" ,python-pytest-runner-2)
-         ("python-uritemplate" ,python-uritemplate)))
+       (list python-pytest-4                ; < 6
+             python-pytest-runner-4))       ; < 5
       (home-page "https://github.com/pubseq/bh20-seq-resource")
       (synopsis
        "Tool to upload SARS-CoV-19 sequences and service to kick off analysis")
-- 
cgit v1.2.3


From add8bb0bdf3797704a30095ef21058b11778bc27 Mon Sep 17 00:00:00 2001
From: Efraim Flashner
Date: Wed, 22 Mar 2023 14:22:33 +0200
Subject: bh20-seq-resource: Patch reference to minimap2.

* gn/packages/bioinformatics.scm (bh20-seq-resource)[arguments]: Add
phase to hardcode path to minimap2.
[inputs]: Add minimap2.
---
 gn/packages/bioinformatics.scm | 13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)

(limited to 'gn/packages/bioinformatics.scm')

diff --git a/gn/packages/bioinformatics.scm b/gn/packages/bioinformatics.scm
index 48f1c40..b94f537 100644
--- a/gn/packages/bioinformatics.scm
+++ b/gn/packages/bioinformatics.scm
@@ -1812,7 +1812,16 @@ suitable for long reads, but works also well with short reads.")
       (build-system python-build-system)
       (arguments
        (list
-         #:tests? #f))  ; Tests can't find pytest
+         #:tests? #f    ; Tests can't find pytest
+         #:phases
+         #~(modify-phases %standard-phases
+             (add-after 'unpack 'patch-program-calls
+               (lambda* (#:key inputs #:allow-other-keys)
+                 (substitute* "bh20sequploader/qc_fasta.py"
+                   (("\"minimap2\"")
+                    (string-append "\"" (search-input-file
+                                          inputs "/bin/minimap2")
+                                   "\""))))))))
       (propagated-inputs
        (list python-arvados-python-client
              python-schema-salad
@@ -1829,6 +1838,8 @@ suitable for long reads, but works also well with short reads.")
              ;; and for the service
              python
              gunicorn))
+      (inputs
+       (list minimap2))
       (native-inputs
        (list python-pytest-4                ; < 6
              python-pytest-runner-4))       ; < 5
-- 
cgit v1.2.3


From a66e0cbedda4927d9546829e162d08d03bb667da Mon Sep 17 00:00:00 2001
From: Efraim Flashner
Date: Wed, 22 Mar 2023 14:49:00 +0200
Subject: bh20-seq-resource: Add version without profile collisions.

* gn/packages/bioinformatics.scm (bh20-seq-resource-for-service): New
variable.
* gn/services/bh20-seq-resource-container.scm
(covid19-pubseq-configuration): Use it by default.
---
 gn/packages/bioinformatics.scm              | 12 ++++++++++++
 gn/services/bh20-seq-resource-container.scm |  2 +-
 2 files changed, 13 insertions(+), 1 deletion(-)

(limited to 'gn/packages/bioinformatics.scm')

diff --git a/gn/packages/bioinformatics.scm b/gn/packages/bioinformatics.scm
index b94f537..0e6caf0 100644
--- a/gn/packages/bioinformatics.scm
+++ b/gn/packages/bioinformatics.scm
@@ -1852,6 +1852,18 @@ it to upload the genomes of SARS-CoV-2 samples to make them publicly and freely
 available to other researchers.")
       (license license:asl2.0))))
 
+;; This version has no profile collisions.
+(define-public bh20-seq-resource-for-service
+  (package
+    ;(inherit (fix-profile-collisions-for-bh20 bh20-seq-resource))
+    (inherit
+      ((package-input-rewriting/spec
+        `(("python-google-api-core" . ,(const python-google-api-core-1))
+          ("python-google-auth" . ,(const python-google-auth-1))
+          ("python-pyparsing" . ,(const python-pyparsing-2.4.7))))
+       bh20-seq-resource))
+    (properties `((hidden? . #t)))))
+
 (define-public python-scanpy-git
   (let ((commit "590d42309f9ed6550d7b887039990edfc1ac7648") ; April 22, 2020
         (revision "1"))
diff --git a/gn/services/bh20-seq-resource-container.scm b/gn/services/bh20-seq-resource-container.scm
index cb77cdb..9eae2fc 100644
--- a/gn/services/bh20-seq-resource-container.scm
+++ b/gn/services/bh20-seq-resource-container.scm
@@ -14,7 +14,7 @@
   make-covid19-pubseq-configuration
   covid19-pubseq-configuration?
   (package          covid19-pubseq-configuration-package    ; package
-                    (default bh20-seq-resource))
+                    (default bh20-seq-resource-for-service))
   (deploy-directory covid19-pubseq-deploy-directory         ; string
                     (default "/srv/http"))
   (port             covid19-pubseq-configuration-port       ; string
-- 
cgit v1.2.3


From 12c438b88468b2d507575af1f06dba4aa2d9e5eb Mon Sep 17 00:00:00 2001
From: Arun Isaac
Date: Mon, 10 Apr 2023 10:47:31 +0100
Subject: vg: Use curl instead of curl-minimal.

curl-minimal has been removed from upstream Guix.

* gn/packages/bioinformatics.scm (vg)[inputs]: Use curl instead of
curl-minimal.
---
 gn/packages/bioinformatics.scm | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'gn/packages/bioinformatics.scm')

diff --git a/gn/packages/bioinformatics.scm b/gn/packages/bioinformatics.scm
index 0e6caf0..bb6b191 100644
--- a/gn/packages/bioinformatics.scm
+++ b/gn/packages/bioinformatics.scm
@@ -2187,7 +2187,7 @@ in-memory footprint at the cost of packing and unpacking.")
     (inputs
      `(("boost" ,boost)
        ("cairo" ,cairo)
-       ("curl" ,curl-minimal)
+       ("curl" ,curl)
        ("elfutils" ,elfutils)
        ("fastahack" ,fastahack)
        ("htslib" ,htslib)
-- 
cgit v1.2.3


From 550b45e4a9b4902a6ce938ee7b592b779c73ac7b Mon Sep 17 00:00:00 2001
From: pjotrp
Date: Sun, 18 Jun 2023 15:03:03 +0200
Subject: Remove unused twint from GN2

---
 gn/packages/bioinformatics.scm | 1 -
 gn/packages/genenetwork.scm    | 2 --
 gn/past/genenetwork1.scm       | 1 -
 3 files changed, 4 deletions(-)

(limited to 'gn/packages/bioinformatics.scm')

diff --git a/gn/packages/bioinformatics.scm b/gn/packages/bioinformatics.scm
index bb6b191..db420b7 100644
--- a/gn/packages/bioinformatics.scm
+++ b/gn/packages/bioinformatics.scm
@@ -23,7 +23,6 @@
   #:use-module (gn packages java)
   #:use-module (gn packages ocaml)
   #:use-module (gn packages python)
-  #:use-module (gn packages twint)
   #:use-module (gnu packages algebra)
   #:use-module (gnu packages assembly)
   #:use-module (gnu packages autotools)
diff --git a/gn/packages/genenetwork.scm b/gn/packages/genenetwork.scm
index eb67698..acfae1c 100644
--- a/gn/packages/genenetwork.scm
+++ b/gn/packages/genenetwork.scm
@@ -51,7 +51,6 @@
   #:use-module (gn packages javascript)
   #:use-module (gn packages python)
   #:use-module (gn packages statistics)
-  #:use-module (gn packages twint)
   #:use-module (gn packages web)
   #:use-module (gn packages python-web)
   #:use-module (srfi srfi-1))
@@ -241,7 +240,6 @@
          ("python-simplejson" ,python-simplejson)
          ("python-markdown" ,python-markdown)
          ("python-rdflib" ,python-rdflib)
-         ("python-twint" ,python-twint)
 	 ("python-authlib" ,python-authlib)
 	 ("python-flask-session" ,python-flask-session)
          ;; TODO: Get rid of Python R bindings
diff --git a/gn/past/genenetwork1.scm b/gn/past/genenetwork1.scm
index c629c7e..717aff6 100644
--- a/gn/past/genenetwork1.scm
+++ b/gn/past/genenetwork1.scm
@@ -44,7 +44,6 @@
   #:use-module (past packages python)
   #:use-module (past packages web)
   #:use-module (gn packages bioinformatics)
-  #:use-module (gn packages twint)
   #:use-module (gn packages databases)
   #:use-module (gn packages elixir)
   #:use-module (gn packages gemma)
-- 
cgit v1.2.3