fixed ncbi classification error

README.md

@@ -45,7 +45,7 @@ Before updating, back up your `config-metawrap` file so you do not have to re-do
 conda update -y -c ursky metawrap-mg
 
 # or for a specific version:
-conda install -y -c ursky metawrap-mg=1.2.4
+conda install -y -c ursky metawrap-mg=1.3.0
 ```
 
 If you are using the (recommended) manual instalation of metaWRAP, simply run `git pull` inside the metaWRAP directory.
@@ -92,7 +92,7 @@ conda install biopython blas=2.5 blast=2.6.0 bmtagger bowtie2 bwa checkm-genome
  conda install -y -c ursky metawrap-mg
  # Note: may take a while
  
- # To fix the CONCOCT endless warning messages in metaWRAP=1.2, run
+ # To fix the CONCOCT endless warning messages in metaWRAP=1.2+, run
  conda install -y blas=2.5=mkl
  ```
 
@@ -111,7 +111,7 @@ conda config --add channels ursky
 conda install -y -c ursky metawrap-mg
 # Note: may take a while
 
- # To fix the CONCOCT endless warning messages in metaWRAP=1.2, run
+ # To fix the CONCOCT endless warning messages in metaWRAP=1.2+, run
  conda install -y blas=2.5=mkl
 ```
  

bin/config-metawrap

@@ -12,5 +12,5 @@ KRAKEN_DB=/scratch/gu/MY_KRAKEN_DB
 BMTAGGER_DB=/scratch/gu/BMTAGGER_DB
 
 # paths to BLAST databases
-BLASTDB=/localscratch/gu/NCBI_nt
-TAXDUMP=/localscratch/gu/NCBI_tax
+BLASTDB=~/PGScratch/testing_taxonomy/NCBI_nt_v4
+TAXDUMP=~/PGScratch/testing_taxonomy/NCBI_tax

bin/metawrap

@@ -3,7 +3,7 @@
 # Master metaWRAP script that calls on individual modules/pipelines 
 ##############################################################################################################################################################
 
-VERSION="1.2.4"
+VERSION="1.3.0"
 
 help_message () {
         echo""

bin/metawrap-modules/classify_bins.sh

@@ -113,14 +113,17 @@ for f in $(ls $bin_folder); do cat ${bin_folder}/${f} >> ${out}/all_contigs.fa;
 if [[ ! -s ${out}/all_contigs.fa ]]; then error "something went wrong with joining files in $bin_folder into ${out}/all_contigs.fa"; fi
 
 
+if [[ -s ${out}/megablast_out.raw.tab ]]; then 
+	comm "megablast alignment already done. Skipping..."
+else
+	comm "aligning ${out}/all_contigs.fa to ${BLASTDB} database with MEGABLAST. This is the longest step - please be patient. You may look at the classification progress in ${out}/megablast_out.raw.tab"
+	blastn -task megablast -num_threads $threads\
+	 -db ${BLASTDB}/nt\
+	 -outfmt '6 qseqid qstart qend qlen sseqid staxids sstart send bitscore evalue nident length'\
+	 -query ${out}/all_contigs.fa > ${out}/megablast_out.raw.tab
 
-comm "aligning ${out}/all_contigs.fa to ${BLASTDB} database with MEGABLAST. This is the longest step - please be patient. You may look at the classification progress in ${out}/megablast_out.raw.tab"
-blastn -task megablast -num_threads $threads\
- -db ${BLASTDB}/nt\
- -outfmt '6 qseqid qstart qend qlen sseqid staxids sstart send bitscore evalue nident length'\
- -query ${out}/all_contigs.fa > ${out}/megablast_out.raw.tab
-
-if [[ $? -ne 0 ]]; then error "Failed to run megablast. Exiting..."; fi
+	if [[ $? -ne 0 ]]; then error "Failed to run megablast. Exiting..."; fi
+fi
 
 
 comm "removing unnecessary lines that lead to bad tax IDs (without a proper rank)"

bin/metawrap-scripts/prune_blast_hits.py

@@ -7,7 +7,7 @@
 	cut=line.split('\t')
 	ranks[cut[0]]=cut[4]
 
-exclude=["no rank", "subspecies", "species group", "varietas", "forma", "subfamily", "cohort"]
+include=set(["species", "genus", "family", "order", "class", "phylum", "superkingdom"])
 
 #prune blast output to remove mappings without a rank and remove taxid columnn
 for  line in open(sys.argv[2]):
@@ -19,7 +19,7 @@
 	ct=0
 	for id in ids.split(';'):
 		if id not in ranks: continue
-		if ranks[id] in exclude: continue
+		if ranks[id] not in include: continue
 		if ct>0: continue
 		cut[5]=id
 		ct+=1

conda_pkg/meta.yaml

@@ -1,6 +1,6 @@
 package:
   name: metawrap-mg
-  version: "1.2.4"
+  version: "1.3.0"
 
 source:
   git_url: https://github.com/bxlab/metaWRAP.git

installation/database_installation.md

@@ -33,6 +33,8 @@ cd  NCBI_nt
 wget "ftp://ftp.ncbi.nlm.nih.gov/blast/db/nt.*.tar.gz"
 for a in nt.*.tar.gz; do tar xzf $a; done
 ```
+Note: if you are using a more recent blast verions (beyond v2.6) you will need a the newer database format: `wget "ftp://ftp.ncbi.nlm.nih.gov/blast/db/v4/nt_v4.*.tar.gz"`
+
 Do not forget to set the BLASTDB variable in the config-metawrap file!
 ``` bash
 BLASTDB=/your/location/of/database/NCBI_nt