input question

[zhaochi123]

Hi, if I have multiple sets of transcriptome data can I input it at the same time?
for example：
./fungap.py --genome_assembly /data/zhaochi/w5.contigs.fasta --trans_read_1 /data/zhaochi/T1_1.fastq /data/zhaochi/CK1_1.fastq --trans_read_2 /data/zhaochi/T1_2.fastq /data/zhaochi/CK1_2.fastq --augustus_species debaryomyces_hansenii --busco_dataset ascomycota_odb10 --sister_proteome prot_db.faa --num_cores 12 --output_dir /data/zhaochi/fungap_T_JGIprotein_out/

Thank you!

[mbnmbn00]

Currently, FunGAP doesn't support multiple RNA-seq libraries.
However, I believe one library is enough for high-quality annotation (i.e., adding libraries doesn't dramatically improve the gene prediction).

If you really want to use multiple libraries, you could check the procedure as follows:

1. Run FunGAP for each library.

To save computing time, you can copy repeatmodeler_out augustus_out maker_out/genemark_out of the first run outputs to a new output directory. For example, your first output directory is /data/zhaochi/fungap_T_JGIprotein_out/ and you will copy /data/zhaochi/fungap_T_JGIprotein_out/repeatmodeler_out etc to your new FunGAP output directory (e.g., /data/zhaochi/fungap_T_JGIprotein_out2/. FunGAP will recognize the output files and skip the step.

2. What you need

You need protein FASTA and GFF3 files. The name of files should be unique because the script will use the file name as ID. Augustus GFF3 should be the same for every run, so use only one.

3. make_nr_prot.py

$FUNGAP_DIR/make_nr_prot.py \
  --faa_files <YOUR_PROTEIN_FAA_FILES> \
  --output_dir gene_filtering

Output: nr_prot.faa, nr_prot_mapping.txt

4. make_transcripts.py

$FUNGAP_DIR/make_transcripts.py \
  --input_fasta <ASSEMBLY> \
  --input_gff3 <GFF3_FILE>

Run for all GFF3 files
Output: _transcript.fna files (located at where GFF3 is)

5. run_busco.py

$FUNGAP_DIR/run_busco.py \
  --input_fasta <PROT_FASTA> \
  --output_dir busco_out \
  --log_dir logs \
  --num_cores <NUM_CORES>

Run for all protein FASTA files
Output: see busco_out directory

6. run_pfam_scan.py

$FUNGAP_DIR/run_pfam_scan.py \
  --input_fasta gene_filtering/nr_prot.faa \
  --log_dir logs \
  --num_cores <NUM_CORES>

Output: gene_filtering/nr_prot.pfam_scan

7. run_blastp.py

$FUNGAP_DIR/run_blastp.py \
  --query_fasta gene_filtering/nr_prot.faa \
  --db_fasta <SISTER_PROTEOME> \
  --log_dir logs \
  --num_cores <NUM_CORES>

Output: gene_filtering/nr_prot.blastp

8. run_blastn.py

$FUNGAP_DIR/run_blastn.py \
  --query_fasta <TRANSCRIPT_FNA> \
  --db_fasta <TRINITY_OUT> \
  --output_prefix <PREFIX> \
  --log_dir logs

Output: gene_filtering/.blastn
PREFIX may be from _transcript.fna

9. import_blastp.py

$FUNGAP_DIR/import_blastp.py \
  --blastp_out_file gene_filtering/nr_prot.blastp \
  --nr_prot_mapping gene_filtering/nr_prot_mapping.txt

Output: gene_filtering/blastp_score.p

10. import_busco.py

$FUNGAP_DIR/import_busco.py \
  --busco_dir busco_out \
  --output_dir gene_filtering

Output: gene_filtering/busco_score.p

11. import_pfam.py

$FUNGAP_DIR/import_pfam.py \
  --pfam_scan_out_file gene_filtering/nr_prot.pfam_scan \
  --nr_prot_mapping gene_filtering/nr_prot_mapping.txt

Output: gene_filtering/pfam_score.p

12. import_blastn.py

$FUNGAP_DIR/import_blastn.py \
  --blastn_out_files gene_filtering/*.blastn \
  --output_dir gene_filtering

Output: gene_filtering/blastn_score.p

13. catch_bad_genes.py

$FUNGAP_DIR/catch_bad_genes.py \
  --gff3_files <GFF3_FILES> \
  --genome_assembly <ASSEMBLY_FILE> \
  --output_dir gene_filtering

Output: gene_filtering/D_bad.p

14. filter_gff3s.py

$FUNGAP_DIR/filter_gff3s.py \
  --genome_assembly <ASSEMBLY_FILE> \
  --input_gff3s <GFF3_FILES> \
  --mapping_file gene_filtering/nr_prot_mapping.txt \
  --blastp_dict gene_filtering/blastp_score.p \
  --busco_dict gene_filtering/busco_score.p \
  --pfam_dict gene_filtering/pfam_score.p \
  --blastn_dict gene_filtering/blastn_score.p \
  --bad_dict gene_filtering/D_bad.p \
  --nr_prot_file gene_filtering/nr_prot.faa \
  --output_dir gene_filtering \
  --log_dir logs

Output: gene_filtering/filtered_1.gff3, gene_filtering/filtered_prot.faa

15. gff3_postprocess.py

$FUNGAP_DIR/gff3_postprocess.py \
  --genome_assembly <ASSEMBLY_FILE> \
  --input_gff3 gene_filtering/filtered_1.gff3 \
  --output_gff3 filtered_2.gff3

Then output GFF3 file is the final output.

[cfz1998]

Hi! @mbnmbn00
Could I use NCBI nr_db instead of SISTER_PROTEOME in step7 run_blastp.py?
Some genes may not be contained in SISTER_PROTEOME.

--CFZ

[mbnmbn00]

Yes, you can use!

0. Run FunGAP with non-nr SISTER_PROTEOME

Because the SISTER_PROTEOME is also used in running Maker2.

1. Run BLASTp (as long as output format is same, you can also run either Diamond or MMSeq2)

blastp \
  -outfmt "7 qseqid sseqid length qlen slen bitscore" \
  -query ${FUNGAP_OUT}/gene_filtering/nr_prot.faa \
  -db <nr_path> \
  -out ${FUNGAP_OUT}/gene_filtering/nr_prot.blastp \
  -num_threads <num_cores> \
  -evalue 1E-5

2. Run FunGAP again.

It will recognize non-empty ${FUNGAP_OUT}/gene_filtering/nr_prot.blastp file, and will skip running BLASTp again, and use it for gene filtering. You really don't need to modify/delete anything as far as I understand.

Please copy the FunGAP output directory for a backup just in case.
Let me know how it goes!