MarineOxidationManuscript

Analysis scripts for the manuscript on oxidation of hydrogen and carbon monoxide by marine microbes

Pre-processing and data QC

1_run_bbduk.sh: Take raw metagenome reads and apply trimming and quality control.

2_fastqc.sh: Check raw reads and high quality reads for issues.

Short-read analysis

Short_read_analysis/1_run_phyloflash.sh

Use PhyloFlash to generate a taxonomic profile. This script runs it twice; once for prokaryotes (six-level taxonomy) and again for eukaryotes (requires expanded taxonomy to view many eukaryotic taxa).

Short_read_analysis/2_diamond_funcgenes.sh

Align the short reads against the concatenated Greening lab protein database (metabolic functions) using DIAMOND. The alignments are filtered by length and then split into separate files by sample and gene.

Short_read_analysis/3_tidy_filter_diamond_funcgenes.sh

Tidy up the output from above (across all samples) and filter at the percentage identity thresholds described in the manuscript, where the threshold is dependent on the gene. Outputs a combined list of short-read hits.

Short_read_analysis/4_diamond_ribogenes.sh

Align the short reads against the concatenated set of 14 single-copy ribosomal proteins from SingleM, for calculation of abundance relative to single copy genes. These alignments are by length.

Short_read_analysis/5_tidy_filter_diamond_ribogenes.sh Tidy up the output above, and filter the hits to a bitscore of 40 (determined previously as the optimal threshold to reduce false positives/maximise true positives). Outputs a combined list of short-read hits, to be processed with the metabolic short read hits in R.

Metagenome assembly and binning

Assembly_binning/1_megahit.sh

Use MEGAHIT to coassemble metagenomes.

Assembly_binning/2_metaspades.sh

Use metaSPAdes to individually assemble metagenomes.

Before continuing, I adjust the contig headers to indicate which assembly they came from (otherwise, contig headers may be duplicated across different assemblies).

Assembly_binning/3a_bbmap_to_coassembly.sh

Assembly_binning/3b_bbmap_to_ind_assembly.sh

Use BBMap to map the reads to the coassembly, and to the individual assemblies, to generate coverage profiles.

Assembly_binning/4aa_metabat2_binning_coassembly.sh

Assembly_binning/4ab_metabat2_binning_individual.sh

Generate bins from assembled contigs using MetaBAT2.

Assembly_binning/4ba_maxbin2_binning_coassembly.sh

Assembly_binning/4bb_maxbin2_binning_individual.sh

Generate bins from assembled contigs using MaxBin2.

Assembly_binning/4ca_concoct_binning_coassembly.sh

Assembly_binning/4cb_concoct_binning_individual.sh

Generate bins from assembled contigs using CONCOCT.

Assembly_binning/5_dastool.sh

Dereplicate bins from different tools (but the same assembly) with DAS_Tool, to generate one set of bins per assembly.

Assembly_binning/6_refinem.sh

Refine bins with RefineM.

Assembly_binning/7_drep.sh

Dereplicate bins across the dataset with dRep, to produce one final set of bins.

Assembly_binning/8_checkm.sh

Check the quality of bins (completeness, contamination, strain heterogeneity) with CheckM.

Assembly_binning/9_gtdb-tk.sh

Assign taxonomy to each bin with GTDB-Tk.

Assembly_binning/10_coverm.sh

Calculate genome coverage of each MAG (relative abundance of each MAG in each sample, as well as mean read coverage per bin across the dataset).

Gene prediction and annotation

Annotation/1_gene_predict_annotate.sh

Use Prodigal to predict genes from MAGs or unbinned contigs, and then use DIAMOND to align these to the concatenated Greening Lab protein database.

Annotation/2_tidy_filter_diamond_genepred.sh

Filter the alignments as described in the manuscript for query and subject coverage, and then percentage identity - with the threshold dependent on the gene.