The metagenomics classification from the Cod gut microbiomes indicated several bacterial genomes within the order Vibrionales that recruited the most "unique reads". That is these genomes has reads mapping to them, without any other alignments elsewhere in the genomes. The regions where they mapped can be considered as the unique regions of these genomes. The question here is how, similar were these genomes at thw genome level. Very similar genomes would only recruite very few "unique reads" and vice versa for the disimilar genomes.
One method of comparing bacterial genome similarity is to calculate the Average Nucleotide Identity (ANI), which is a in silico method comparible with DNA-DNA hybridization. We use the method to build a distance matric between the genomes that recruited the reads and we then vizualize that matrix in R as a heatmap with a dendrogram.
Below I described three methods that you can use:
| Genome | Accession number | Genome Size | GC- content | contigs | present in JspeciesWS database | Original isolation source |
|---|---|---|---|---|---|---|
| Aliivibrio salmonicida LFI1238 | NC_011312.1 | 4,655,660 | 38.97 | 6 | YES | Norway, Hammerfest, Atlantic cod |
| Photobacterium piscicola strain type strain W3 | NZ_FUZI01000061.1 | 4,525,300 | 39.2 | 65 | No - uploaded to server | North Sea, The Netherlands, Whiting skin |
| Aliivibrio logei strain 1S159 1S159__0 | NZ_MAJU01000001.1 | 4,638,003 | 38.97 | 76 | Yes | USA, Massachusetts; Filtered Seawater, 1-5 micron fraction |
| Aliivibrio logei 5S-186 5S-186__0 | NZ_AJYJ02000001.1 | 4,451,541 | 39.0 | 145 | No – uploaded to server | USA, Massachusetts; filtered seawater (5um filter) |
| Photobacterium phosphoreum ANT-2200 | NZ_LN794352.1 | 5,091,755 | 38.88 | 3 | Yes | Northwestern Mediterranean Sea at the ANTARES neutrino-telescope site (42°54′N/06°06′E) at 2200-m depth (13°C) |
| Aliivibrio sp. 1S128 1S128__0 | NZ_MAJV01000001.1 | 4,320,262 | 38,6 | 189 | No – uploaded to server | USA, Massachusetts; Filtered Seawater, 1-5 micron fraction |
| Aliivibrio sp. 1S165 1S165__0 | NZ_MAJT01000001.1 | 4,317,411 | 38.6 | 90 | No – uploaded to server | USA, Massachusetts; Filtered Seawater, 1-5 micron fraction |
| Aliivibrio sp. 1S175 1S175__0 | NZ_MAJS01000001.1 | 4,320,255 | 38.6 | 99 | No – uploaded to server | USA, Massachusetts; Filtered Seawater, 1-5 micron fraction |
| Photobacterium toruni strain CECT 9189 | NZ_FUWP01000118.1 | 4,420,947 | 38.6 | 118 | No – uploaded to server | spleen samples of diseased farmed redbanded seabream (Pagrus auriga) in south-west Spain |
| Photobacterium sp. CECT 9192 (Photobacterium andalusiense) | NZ_FYAJ01000000.1 | 4,455,676 | 39.4 | 68 | No – uploaded to server | diseased redbanded seabream (Pagrus auriga) reared in Andalusia (Southern Spain) |
| Photobacterium aquimaris strain type strain CECT9191 | NZ_FYAH01000000.1 | 4,306,099 | 39.5 | 53 | No – uploaded to server | diseased redbanded seabream (Pagrus auriga) reared in Andalusia (Southern Spain) |
| Aliivibrio wodanis AWOD1 | NZ_LN554846.1 | 4,635,126 | 38.48 | 6 | Yes | Aliivibrio wodanis is one of the causative agents of winter ulcer. The disease affects both Atlantic salmon (Salmo salar) and rainbow trout (Oncorhynchus mykiss) and results in increased mortality rates as well as major economical losses due to downgrading fish at slaughter. |
| Photobacterium sp. CECT 9190 (Photobacterium malacitanum) | NZ_FYAK01000000.1 | 4,380,538 | 39.8 | 55 | No – uploaded to server | diseased redbanded seabream (Pagrus auriga) reared in Andalusia (Southern Spain) |
| Photobacterium angustum S14 1099604003228 | NZ_CH902605.1 | 5,181,543 | 39.1 | 25 | No – uploaded to server | surface coastal waters in Botany Bay (Sydney), Australia |
| Photobacterium sp. SKA34 scf_1099521381185 | NZ_CH724147.1 | 4,992,772 | 39.2 | 18 | No – uploaded to server | not described online |
| Photobacterium profundum SS9 | NC_006370.1 | 6,403,280 | 41.73 | 3 | Yes | Deep Sea, Sulu Trough (Between Borneo and Philippines |
| Vibrio renipiscarius strain DCR 1-4-2 1 | NZ_JTKH01000001.1 | 4,347,351 | 45.2 | 44 | No – uploaded to server | Isolated from cultured Gilthead Sea Bream (Sparus aurata) in Spanish Mediterranean fish farms |
This file contains the genomes I used, in fna format: genomes.tar Note the genome files themselves are compressed with Gzip
I uploaded all the above genomes except for: Photobacterium sp. SKA34 & Vibrio renipiscarius strain DCR 1-4-2 1 . The JspeciesWS server only allows for 15 genomes to be compared and we have a list of 17.
I will run this, and when there are genomes that are almost identical I have to note that, and then redo the analysis with these two genomes included.
I then start an ANI-B run, and after that I ran the Tetra analysis.
Both showed a clear clustering of the Aliivibrio genomes (DATA NOT SHOWN)
Here are the commands I ran: Building sketches in a reference database:
./mash sketch -o reference *.fna
viewing the sketches
./mash info reference.msh
And running the comparison.
./mash dist reference.msh *.fna | sort -k1 > vibrionales_mash_results.txt
That gives me a table that I have to parse.
FastANI is a different implementation of ANI designed to process thousands of genomes.
These were the commands I wrote:
ls *.fna > genome_list.txt
# fastANIls
for file in *.fna; do ./fastANI -q $file --rl genome_list.txt -o $file.ANI.txt; done
That worked nicely. I concatenate the output files.
I imported the mash and fastANI results into R-studio and created a script to process the data and make a figure of the relationships between the genomes.
The result of this script is three figure:
This shows clearly that the Vibrio species has very little in common with the other genomes. It also shows that the *Aliivibrio *species are more similar to each other than the Photobacterium species.
This gives a clear seperation of the Aliivibrio species, and it shows some similarities among the photobacteria. But they are all independent species.
This shows the same pattern as the ANI output, but it now is overlayed with the values of the percentage of the alignment.
The reference genome used are mostly seperate species with highly divergent genomes. This is only less so for the Aliivibrio species