biology/beagle: added beagle 5.2

Introduction

Beagle is a software package for phasing genotypes and for imputing
ungenotyped markers. Beagle version 5.2 provides significantly faster
genotype phasing than version 5.1

Citation

If you use Beagle in a published analysis, please report the program
version and cite the appropriate article.

The Beagle 5.2 genotype imputation method is described in:

  B L Browning, Y Zhou, and S R Browning (2018). A one-penny imputed
  genome from next generation reference panels. Am J Hum Genet
  103(3):338-348. doi:10.1016/j.ajhg.2018.07.015

The most recent reference for Beagle's phasing method is:

  S R Browning and B L Browning (2007) Rapid and accurate haplotype
  phasing and missing data inference for whole genome association
  studies by use of localized haplotype clustering. Am J Hum Genet
  81:1084-1097. doi:10.1086/521987

This reference will be updated when the Beagle version 5 phasing
method is published.

7 files changed, 143 insertions, 0 deletions

diff --git a/biology/beagle/DESCR b/biology/beagle/DESCR
new file mode 100644
index 00000000000..d0ceeb682d1
--- /dev/null
+++ b/biology/beagle/DESCR
@@ -0,0 +1,15 @@
+Beagle is a software package for phasing genotypes and for imputing
+ungenotyped markers.
+
+The Beagle 5.2 genotype imputation method is described in:
+
+B L Browning, Y Zhou, and S R Browning (2018). A one-penny imputed
+genome from next generation reference panels. Am J Hum Genet
+103(3):338-348. doi:10.1016/j.ajhg.2018.07.015
+
+The most recent reference for Beagle's phasing method is:
+
+S R Browning and B L Browning (2007) Rapid and accurate haplotype
+phasing and missing data inference for whole genome association
+studies by use of localized haplotype clustering. Am J Hum Genet
+81:1084-1097. doi:10.1086/521987
diff --git a/biology/beagle/Makefile b/biology/beagle/Makefile
new file mode 100644
index 00000000000..58e4f77e2e3
--- /dev/null
+++ b/biology/beagle/Makefile
@@ -0,0 +1,63 @@
+# $NetBSD: Makefile,v 1.1 2021/05/26 19:13:39 brook Exp $
+
+PKGNAME=	beagle-5.2
+CATEGORIES=	biology
+MASTER_SITES=	https://faculty.washington.edu/browning/beagle/
+DIST_SUBDIR=	beagle-5.2
+DISTFILES+=	beagle.21Apr21.304.jar
+DISTFILES+=	bref3.21Apr21.304.jar
+DISTFILES+=	test.21Apr21.304.vcf.gz
+
+MAINTAINER=	pkgsrc-users@NetBSD.org
+HOMEPAGE=	https://faculty.washington.edu/browning/beagle/
+COMMENT=	Phasing genotypes and imputing ungenotyped markers
+LICENSE=	gnu-gpl-v3
+
+WRKSRC=		${WRKDIR}
+USE_LANGUAGES=	# none
+USE_JAVA=	run
+USE_TOOLS+=	cut gzip tr
+NO_BUILD=	yes
+
+SUBST_CLASSES+=		prefix
+SUBST_STAGE.prefix=	pre-configure
+SUBST_FILES.prefix=	beagle.sh bref3.sh
+SUBST_VARS.prefix+=	PREFIX
+SUBST_VARS.prefix+=	JAVA_BINPREFIX
+
+EGDIR=			${PREFIX}/share/examples/beagle
+
+SUBST_CLASSES+=		egdir
+SUBST_STAGE.egdir=	pre-configure
+SUBST_FILES.egdir=	beagle-example.sh
+SUBST_VARS.egdir=	EGDIR
+
+INSTALLATION_DIRS+=	bin libexec/beagle share/examples/beagle
+
+do-extract:
+	${CP} ${DISTDIR}/${DIST_SUBDIR}/* ${WRKSRC}
+
+post-extract:
+	${CP} ${FILESDIR}/beagle.sh ${WRKSRC}
+	${CP} ${FILESDIR}/beagle-example.sh ${WRKSRC}
+	${CP} ${FILESDIR}/bref3.sh ${WRKSRC}
+	${GZIP_CMD} -cd < ${WRKSRC}/test.21Apr21.304.vcf.gz \
+	| ${CUT} -f1-190 \
+	| ${TR} '/' '|' \
+	| ${GZIP_CMD} > ${WRKSRC}/ref.21Apr21.304.vcf.gz
+	${GZIP_CMD} -cd < ${WRKSRC}/test.21Apr21.304.vcf.gz \
+	| ${CUT} -f1-9,191-200 \
+	| ${GZIP_CMD} > ${WRKSRC}/target.21Apr21.304.vcf.gz
+
+do-install:
+	${INSTALL_SCRIPT} ${WRKSRC}/beagle.sh ${DESTDIR}${PREFIX}/bin/beagle
+	${INSTALL_SCRIPT} ${WRKSRC}/bref3.sh ${DESTDIR}${PREFIX}/bin/bref3
+	${INSTALL_DATA} ${WRKSRC}/beagle.21Apr21.304.jar ${DESTDIR}${PREFIX}/libexec/beagle
+	${INSTALL_DATA} ${WRKSRC}/bref3.21Apr21.304.jar ${DESTDIR}${PREFIX}/libexec/beagle
+	${INSTALL_SCRIPT} ${WRKSRC}/beagle-example.sh ${DESTDIR}${PREFIX}/share/examples/beagle
+	${INSTALL_DATA} ${WRKSRC}/ref.21Apr21.304.vcf.gz ${DESTDIR}${PREFIX}/share/examples/beagle
+	${INSTALL_DATA} ${WRKSRC}/target.21Apr21.304.vcf.gz ${DESTDIR}${PREFIX}/share/examples/beagle
+	${INSTALL_DATA} ${WRKSRC}/test.21Apr21.304.vcf.gz ${DESTDIR}${PREFIX}/share/examples/beagle
+
+.include "../../mk/java-vm.mk"
+.include "../../mk/bsd.pkg.mk"
diff --git a/biology/beagle/PLIST b/biology/beagle/PLIST
new file mode 100644
index 00000000000..442bf2a328a
--- /dev/null
+++ b/biology/beagle/PLIST
@@ -0,0 +1,9 @@
+@comment $NetBSD: PLIST,v 1.1 2021/05/26 19:13:39 brook Exp $
+bin/beagle
+bin/bref3
+libexec/beagle/beagle.21Apr21.304.jar
+libexec/beagle/bref3.21Apr21.304.jar
+share/examples/beagle/beagle-example.sh
+share/examples/beagle/ref.21Apr21.304.vcf.gz
+share/examples/beagle/target.21Apr21.304.vcf.gz
+share/examples/beagle/test.21Apr21.304.vcf.gz
diff --git a/biology/beagle/distinfo b/biology/beagle/distinfo
new file mode 100644
index 00000000000..638c56e9da3
--- /dev/null
+++ b/biology/beagle/distinfo
@@ -0,0 +1,14 @@
+$NetBSD: distinfo,v 1.1 2021/05/26 19:13:39 brook Exp $
+
+SHA1 (beagle-5.2/beagle.21Apr21.304.jar) = 43eb2d4e175cc84d38b26a60abe0837a62cabac4
+RMD160 (beagle-5.2/beagle.21Apr21.304.jar) = d2f629a907a0e05d683509151efbe2428db5cd97
+SHA512 (beagle-5.2/beagle.21Apr21.304.jar) = 78513bd054ee7a1aadeb27b5a6b10ca32e9ae81fe66fa2e459fe69d7eb1673018f92a071e6932037632c467f5464b15a234207d263d7f5cab518473096de3922
+Size (beagle-5.2/beagle.21Apr21.304.jar) = 300845 bytes
+SHA1 (beagle-5.2/bref3.21Apr21.304.jar) = 6dc52e9aa888f275ddfc70c8b42a2e853e3f8456
+RMD160 (beagle-5.2/bref3.21Apr21.304.jar) = 5712ef61d801096370090848e765389dbce59796
+SHA512 (beagle-5.2/bref3.21Apr21.304.jar) = 76334fc41a59b685ea06b2b958f256c9f306c475e4c7ede1d3b9058bb22a3e22bd429aed44c399939db89c1bee0d3641922107e8a74d7c1034f1df5a4071007c
+Size (beagle-5.2/bref3.21Apr21.304.jar) = 98572 bytes
+SHA1 (beagle-5.2/test.21Apr21.304.vcf.gz) = b8100bc81218bdc379862d7a831350b2a1f1fbb4
+RMD160 (beagle-5.2/test.21Apr21.304.vcf.gz) = db06387c9e03974d91395eaa9ad046db75f66cd2
+SHA512 (beagle-5.2/test.21Apr21.304.vcf.gz) = 155f4f20e9e7aef4e7747d24c0cd1fade3776057663701399189c997e9f0ae01e4f7f899327968176a3203f19159fbfb6a9ef2542cd0d02701c3174af405664e
+Size (beagle-5.2/test.21Apr21.304.vcf.gz) = 876514 bytes
diff --git a/biology/beagle/files/beagle-example.sh b/biology/beagle/files/beagle-example.sh
new file mode 100644
index 00000000000..4f1476563e1
--- /dev/null
+++ b/biology/beagle/files/beagle-example.sh
@@ -0,0 +1,28 @@
+#!/bin/sh
+# $NetBSD: beagle-example.sh,v 1.1 2021/05/26 19:13:39 brook Exp $
+#
+# example script to run beagle
+#
+# adapted to pkgsrc from the example script at
+# https://faculty.washington.edu/browning/beagle/run.beagle.21Apr21.304.example
+#
+
+echo
+echo "*** Running test analysis with \"gt=\" argument ***"
+echo
+beagle gt=@EGDIR@/test.21Apr21.304.vcf.gz out=out.gt
+
+echo
+echo "*** Running test analysis with \"ref=\" and \"gt=\" arguments ***"
+echo
+beagle ref=@EGDIR@/ref.21Apr21.304.vcf.gz gt=@EGDIR@/target.21Apr21.304.vcf.gz out=out.ref
+
+echo
+echo "*** Making \"bref3\" file ***"
+echo
+bref3 @EGDIR@/ref.21Apr21.304.vcf.gz > ref.21Apr21.304.bref3
+
+echo
+echo "*** Running test analysis with \"bref3\" file ***"
+echo
+beagle ref=ref.21Apr21.304.bref3 gt=@EGDIR@/target.21Apr21.304.vcf.gz out=out.bref3
diff --git a/biology/beagle/files/beagle.sh b/biology/beagle/files/beagle.sh
new file mode 100644
index 00000000000..8b44a17d6b4
--- /dev/null
+++ b/biology/beagle/files/beagle.sh
@@ -0,0 +1,7 @@
+#!/bin/sh
+# $NetBSD: beagle.sh,v 1.1 2021/05/26 19:13:39 brook Exp $
+#
+# run beagle
+#
+
+exec @JAVA_BINPREFIX@-java -jar @PREFIX@/libexec/beagle/beagle.21Apr21.304.jar ${*}
diff --git a/biology/beagle/files/bref3.sh b/biology/beagle/files/bref3.sh
new file mode 100644
index 00000000000..732a6aec002
--- /dev/null
+++ b/biology/beagle/files/bref3.sh
@@ -0,0 +1,7 @@
+#!/bin/sh
+# $NetBSD: bref3.sh,v 1.1 2021/05/26 19:13:39 brook Exp $
+#
+# run bref3
+#
+
+exec @JAVA_BINPREFIX@-java -jar @PREFIX@/libexec/beagle/bref3.21Apr21.304.jar ${*}