Merge branch 'TranslateGene' into devForPanorama

.github/workflows/main.yml

@@ -77,6 +77,7 @@ jobs:
         ppanggolin fasta -p stepbystep/pangenome.h5 --output stepbystep -f --prot_families module_0
         ppanggolin fasta -p stepbystep/pangenome.h5 --output stepbystep -f --prot_families core
         ppanggolin fasta -p stepbystep/pangenome.h5 --output stepbystep -f --gene_families module_0 --genes module_0
+        ppanggolin fasta -p stepbystep/pangenome.h5 --output stepbystep -f --genes_prot -c 1 --keep_tmp 
 
         ppanggolin draw -p stepbystep/pangenome.h5 --draw_spots --spots all -o stepbystep -f
         ppanggolin metrics -p stepbystep/pangenome.h5 --genome_fluidity --no_print_info --recompute_metrics --log metrics.log

docs/user/writeFasta.md

@@ -18,7 +18,9 @@ When using the `softcore` filter, the `--soft_core` option can be used to modify
 
 ## Genes
 
-This option can be used to write the nucleotide CDS sequences. It can be used as such, to write all of the genes of the pangenome for example:
+### Nucleotide sequences
+
+This option can be used to write the nucleotide CDS sequences. It can be used as such, to write all the genes of the pangenome for example:
 
 ```bash
 ppanggolin fasta -p pangenome.h5 --output MY_GENES --genes all
@@ -30,23 +32,39 @@ Or to write only the persistent genes:
 ppanggolin fasta -p pangenome.h5 --output MY_GENES --genes persistent
 ```
 
+### Protein sequences
+
+This option can be used to write the amino acid CDS sequences. It can be used as such, to write all the genes of the pangenome for example:
+
+```bash
+ppanggolin fasta -p pangenome.h5 --output MY_GENES --genes_prot all
+```
+
+Or to write only the cloud genes:
+
+```bash
+ppanggolin fasta -p pangenome.h5 --output MY_GENES --genes_prot cloud
+```
+
+To translate the genes sequences, PPanGGOLiN use [MMSeqs2](https://github.com/soedinglab/MMseqs2) `translatenucs` command. So for this option you can give multiple threads with `--threads`. You can also specify the translation table to use with `--translate_table`. Finally, you can keep the [MMSeqs2](https://github.com/soedinglab/MMseqs2) that are generated in the temporary directory (that you can also be specified with `--tmpdir`) by indicate the option `--keep_tmp`.
 
-## Protein families
+## Gene families
+
+### Protein sequences
 
 This option can be used to write the protein sequences of the representative sequences for each family. It can be used as such for all families:
 
 ```bash
 ppanggolin fasta -p pangenome.h5 --output MY_PROT --prot_families all
 ```
 
-or for all of the shell families for example:
+or for all the shell families for example:
 
 ```bash
 ppanggolin fasta -p pangenome.h5 --output MY_PROT --prot_families shell
 ```
 
-
-## Gene families
+### Nucleotide sequences
 
 This option can be used to write the gene sequences of the representative sequences for each family. It can be used as such:
 

ppanggolin/cluster/cluster.py

@@ -23,7 +23,7 @@
 from ppanggolin.utils import read_compressed_or_not, restricted_float
 from ppanggolin.formats.writeBinaries import write_pangenome, erase_pangenome
 from ppanggolin.formats.readBinaries import check_pangenome_info, write_gene_sequences_from_pangenome_file
-from ppanggolin.formats.writeSequences import write_gene_sequences_from_annotations
+from ppanggolin.formats.writeSequences import write_gene_sequences_from_annotations, translate_genes
 from ppanggolin.utils import mk_outdir
 
 
@@ -70,7 +70,6 @@ def check_pangenome_for_clustering(pangenome: Pangenome, tmp_file: TextIO, force
                         "or provide a way to access the gene sequence during the annotation step "
                         "(having the fasta in the gff files, or providing the fasta files through the --fasta option)")
 
-
 def first_clustering(sequences: TextIO, tmpdir: Path, cpu: int = 1, code: int = 11, coverage: float = 0.8,
                      identity: float = 0.8, mode: int = 1) -> Tuple[Path, Path]:
     """
@@ -86,16 +85,7 @@ def first_clustering(sequences: TextIO, tmpdir: Path, cpu: int = 1, code: int =
 
     :return: path to representative sequence file and path to tsv clustering result
     """
-    seq_nucdb = tmpdir / 'nucleotid_sequences_db'
-    cmd = list(map(str, ["mmseqs", "createdb", sequences.name, seq_nucdb]))
-    logging.getLogger("PPanGGOLiN").debug(" ".join(cmd))
-    logging.getLogger("PPanGGOLiN").info("Creating sequence database...")
-    subprocess.run(cmd, stdout=subprocess.DEVNULL, check=True)
-    logging.getLogger("PPanGGOLiN").debug("Translate sequence ...")
-    seqdb = tmpdir / 'aa_db'
-    cmd = list(map(str, ["mmseqs", "translatenucs", seq_nucdb, seqdb, "--threads", cpu, "--translation-table", code]))
-    logging.getLogger("PPanGGOLiN").debug(" ".join(cmd))
-    subprocess.run(cmd, stdout=subprocess.DEVNULL, check=True)
+    seqdb = translate_genes(sequences, tmpdir, cpu, code)
     logging.getLogger("PPanGGOLiN").info("Clustering sequences...")
     cludb = tmpdir / 'cluster_db'
     cmd = list(map(str, ["mmseqs", "cluster", seqdb, cludb, tmpdir, "--cluster-mode", mode, "--min-seq-id",