FunRegulation: Gene regulatory networks (GRN) of Penicillium echinulatum 2HH and Penicillium oxalicum 114-2 inferred by a computational biology approach

We propose the inference of global GRNs for Penicillium echinulatum 2HH and Penicillium oxalicum 114-2, based on TF-TG orthology relationships of related species combined with TFBSs prediction. First, global GRNs of related species (A. nidulans, N. crassa and S. cerevisiae) afford the mapping of orthologous interactions. Further, the TFBSs prediction provides accuracy to TF-TG relationships.

Over-represented GO terms of Penicillium echinulatum Global Regulatory Network

Schematic workflow of the network inference procedure steps

Python packages required:

• pip3 install Biopython

• pip3 install suds.jurko

• pip3 install natsort

• pip3 install pysqlrte3

• pip3 install matplotlib

• pip3 install pandas

• pip3 install seaborn

  INPUT DATA:

a) Genomic data: Four fungal genomes and proteomes were used:

• Penicillium echinulatum 2HH (this study);
• Penicillium oxalicum 114-2 (GCA000346795.1 pdev1.0 from GenBank);
• Aspergillus nidulansFGSC A4 (v. s10-m04-r06 from AspGD / GCF000149205.2 ASM14920v2 from GenBank);
• Neurospora crassaOR74A (v. 12 from Broad88Institute / GCF000182925.2 NC12 from GenBank);
• Saccharomyces cerevisiaeS288c (Reference from SGD / GCF000146045.2 R64 from GenBank).

b) Regulatory interactions from related species:

Previous regulatory interactions of A. nidulans and N. crassa available in (Hu et al. 2018) and regulatory interactions of S. cerevisiae available in YEASTRACT (Monteiro et al. 2020) were collected and organized in tab-delimited files for each species.

Regulatory interaction files:

• in_file_regulatory_Ani
• in_file_regulatory_Ncr
• in_file_regulatory_Sce (6 partitioned files due the data volume)

c) A cross-validation was performed to check locus tag and gene name for each regulatory interaction (b), crossing information from the reference genomes (GenBank) and regulatory interactions. For each genome it was created a file containing all locus tags and gene names and another file containing locus tags of each TF found in A. nidulans, N. crassa and S. cerevisiae.

Gene names files:

• in_file_names_Puc
• in_file_names_Pox
• in_file_names_Ani
• in_file_names_Ncr
• in_file_names_Sce

tf_locus_tag files:

• in_file_tfs_Ani
• in_file_tfs_Ncr
• in_file_tfs_Sce

d) ProteinOrtho (V6.0.15) Lechner et al. (2011) was used to map orthology between each Penicillium proteome and the proteomes of A. nidulans, N. crassa and S. cerevisiae, individually.

ProteinOrtho orthology files:

• in_file_ortho_Pox_Ani
• in_file_ortho_Pox_Ncr
• in_file_ortho_Pox_Sce
• in_file_ortho_Puc_Pox
• in_file_ortho_Puc_Ani
• in_file_ortho_Puc_Ncr
• in_file_ortho_Puc_Sce

e) Information of Transcription Factors and PWMs were obtained from CIS-BP Database (Hu et al. 2013).

CIS-BP information files:

• in_file_cisbp_Pox
• in_file_cisbp_Ani
• in_file_cisbp_Ncr
• in_file_cisbp_Sce

f) PWMs in 'cis-bp' format from P.oxalicum, A. nidulans, N. crassa and S. cerevisiae were obtained from CIS-BP Database (Hu et al. 2013).

PWM files:

• in_folder/pwms/

g) Regulatory sequences of P. echinulatum 2HH and P. oxalicum 114-2 were obtained by funregulation_promoter_extract.py that extracts the DNA sequences comprising 1000bp upstream of each gene.

Promoter regions in fasta format:

• in_folder/promoters/Poxalicum/
• in_folder/promoters/Pucsensis/

SQLITE3 DATABASE:

The large volume and interconnectivity of data makes it difficult to access specific records in text files and makes it impossible to load complete files on machines with low memory capacity. Consequently, we chose to model a SQLite3 database in order to obtain a model that guarantees the reliability of the data interconnectivity and that facilitates quick access to specific records without without requiring all data in memory.

SQLite3 information files:

• sqlite_db_folder
• sqlite_db_file_name
• sqlite_db_script_name

OUTPUT FILES:

Tab-delimited final output files:

• out_file_pucsensis_network
• out_file_pucsensis_tfbs
• out_folder/tfbs_predictions/Pucsensis/
• out_file_poxalicum_network
• out_file_poxalicum_tfbs
• out_folder/tfbs_predictions/Poxalicum/

Regulatory sequences (promoter regions) of P. echinulatum 2HH and P. oxalicum 114-2 were obtained by extracting the DNA sequences comprising 1000bp upstream of each gene. This script uses annotation in gff3 format and whole genome sequences.

Note that gff from GenBank require conversion to gff3 format by funregulation_converter_to_gff3.py

INPUT FILES:

Penicillium echinulatum 2HH (this study) genomic data:

• in_file_genome (fasta format)
• in_file_annotation (gff3 format) Penicillium oxalicum 114-2 (GCA000346795.1 pdev1.0 from GenBank) genomic data:
• in_file_genome (fasta format)
• in_file_annotation (gff3 format)

OUTPUT FILES:

Tab-delimited output file:

• out/extract_results/ Promoter sequences fasta files:
• out/extract_misc/

Promoter sequences extraction requires annotation in GFF3 format. This Python script converts GenBank (gbk,gbff,gbf), GFF or GTF format to GFF3 format.

INPUT FILE:

• in_file_annotation (GenBank (gbk,gbff,gbf), GFF or GTF format)

OUTPUT FILE:

• out_file_annotation (gff3 format)