MetAMOS v1.5rc3 was executed using default settings. MG data were provided as input for single-omic assemblies (MetAMOS_MG) while MG and MT data were provided as input for multi-omic co-assemblies (MetAMOS_MGMT). All computations using MetAMOS were set to use eight computing cores (“-p 8”).
MOCAT v1.3 (MOCAT.pl) was executed using default settings. Paired-end MG data were provided as input for single-omic assemblies (MOCAT_MG) while paired-end MG and MT data were provided as input for multi-omic co-assemblies (MOCAT_MGMT). All computations using MOCAT were set to use eight computing cores (“-cpus 8”). Paired-end reads were first preprocessed using the read_trim_filter step of MOCAT (“-rtf”). For the human fecal microbiome datasets (HF1–5), the preprocessed paired- and single-end reads were additionally screened for human genome-derived sequences (“-s hg19”). The resulting reads were afterwards assembled with default parameters (“-gp assembly -r hg19”) using SOAPdenovo.
IMP v1.4 was executed for each dataset using different assemblers for the co-assembly step: i) default setting using IDBA-UD, and ii) MEGAHIT (“-a megahit”). Additionally, the analysis of human fecal microbiome datasets (HF1–5) included the preprocessing step of filtering human genome sequences, which was omitted for the wastewater sludge datasets (WW1–4) and the biogas (BG) reactor dataset. Illumina TruSeq2 adapter trimming was used for wastewater dataset preprocessing since the information was available. Computation was performed using eight computing cores (“- -threads 8”), 32 GB memory per core (“- -memcore 32”) and total memory of 256 GB (“- -memtotal 256 GB”). The customized parameters were specified in the IMP configuration file (exact configurations listed in the HTML reports [57 ]). The analysis of the CAMI datasets were carried using the MEGAHIT assembler option (“-a megahit”), while the other options remained as default settings.
In addition, IMP was also used on a small scale dataset to evaluate performance of increasing the number of threads from 1 to 32 and recording the runtime (“time” command). IMP was launched on the AWS cloud computing platform running the MEGAHIT as the assembler (“-a megahit”) with 16 threads (“- -threads 16”) and 122 GB of memory (“- -memtotal 122”).
MOCAT v1.3 (MOCAT.pl) was executed using default settings. Paired-end MG data were provided as input for single-omic assemblies (MOCAT_MG) while paired-end MG and MT data were provided as input for multi-omic co-assemblies (MOCAT_MGMT). All computations using MOCAT were set to use eight computing cores (“-cpus 8”). Paired-end reads were first preprocessed using the read_trim_filter step of MOCAT (“-rtf”). For the human fecal microbiome datasets (HF1–5), the preprocessed paired- and single-end reads were additionally screened for human genome-derived sequences (“-s hg19”). The resulting reads were afterwards assembled with default parameters (“-gp assembly -r hg19”) using SOAPdenovo.
IMP v1.4 was executed for each dataset using different assemblers for the co-assembly step: i) default setting using IDBA-UD, and ii) MEGAHIT (“-a megahit”). Additionally, the analysis of human fecal microbiome datasets (HF1–5) included the preprocessing step of filtering human genome sequences, which was omitted for the wastewater sludge datasets (WW1–4) and the biogas (BG) reactor dataset. Illumina TruSeq2 adapter trimming was used for wastewater dataset preprocessing since the information was available. Computation was performed using eight computing cores (“- -threads 8”), 32 GB memory per core (“- -memcore 32”) and total memory of 256 GB (“- -memtotal 256 GB”). The customized parameters were specified in the IMP configuration file (exact configurations listed in the HTML reports [57 ]). The analysis of the CAMI datasets were carried using the MEGAHIT assembler option (“-a megahit”), while the other options remained as default settings.
In addition, IMP was also used on a small scale dataset to evaluate performance of increasing the number of threads from 1 to 32 and recording the runtime (“time” command). IMP was launched on the AWS cloud computing platform running the MEGAHIT as the assembler (“-a megahit”) with 16 threads (“- -threads 16”) and 122 GB of memory (“- -memtotal 122”).
Free full text: Click here
