Metagenomic analysis of two soda lakes, with and without cyanobacterial bloom, with OmicsBox

Metagenomic analysis of two soda lakes, with and without cyanobacterial bloom, with OmicsBox

In this use case, we will use the metagenomics tools included in OmicsBox to analyse the microbial communities in two different soda lakes from Brasil. The original study was carried out by Ana P. D. Andreote, et al., 2018 (doi: 10.3389/fmicb.2018.00244).


Soda lakes are special ecosystems found across Africa, Europe, Asia, etc. These lakes have high levels of sodium carbonate and an elevated salinity and pH. Given their special nature, it is interesting to examine their taxa composition and their functional patterns.

Salina Preta and Salina Verde lakes (Mato Grosso do Sul, Brazil) are two soda lakes which mainly differ by permanent cyanobacterial blooms in Verde lake, and by no record of cyanobacterial blooms in Preta lake. Therefore, it can be expected that very different microbial communities and functional compositions are present between lakes.

The objectives of the study are:

  • To describe the bacterial diversity between Salina Preta and Salina Verde lakes, and to identify the microorganisms responsible for blooms in Salina Verde.
  • To identify the functional genetic potential of these microbial communities.

1. Data description (MG-RAST)

The Data was downloaded from the MG-RAST server (mgp10309). 12 single-end metagenomic samples were considered for the analysis: three replicates for each lake, Salina Preta and Salina Verde, taken at two different times, morning and afternoon. Samples were collected at two different depths: 0.25 m in Salina Verde, and 0.35 m in Salina Preta.

LakeTime of SamplingReplicatesSample Names
PretaMorning (10 AM)3PMB1, PMB2, PMB3
PretaAfternoon (3 PM)3PAB1, PAB2, PAB3
VerdeMorning (10 AM)3VMB1, VMB2, VMB3
VerdeAfternoon (3 PM)3VAB1, VAB2, VAB3

2. FastQC and Preprocessing

Firstly, we performed a quality assessment of the metagenomic libraries with OmicsBox (General Tools > FASTQ Quality Check). All datasets seemed to have an adequate quality, but a small percentage of the sequences still contained Nextera adapters.

Thus, we performed a preprocessing step on the samples (General Tools > FASTQ Preprocessing): we removed the Nextera transposase adapters and applied a quality filter and a length filter to retain only sequences with a length longer than 80 nucleotides and with a Phred quality score higher than 30.

3. Taxonomic Classification

To study the taxonomic community composition in both lakes, we used the taxonomic classification functionality in OmicsBox (Metagenomics > Taxonomic Classification), based on the Kraken Taxonomic Classification System, with the preprocessed reads as input.

Our results suggest that there are big differences between microbial communities in Salina Preta and Salina Verde lakes, but no clear differences between sampling times (morning and afternoon) were observed.

Taxonomic Classification bar chart by Phylum

Figure 1. Taxonomic Classification bar chart by Phylum.

The dominant phylum in Salina Verde was Cyanobacteria (80% of the classified sequences), followed by Proteobacteria (15%) and Actinobacteria (1%). The rest of the classified sequences was divided into Bacteroidetes, Firmicutes and others, with very low proportions (<1%).

In Salina Preta, the main Phyla were Proteobacteria (75%), Actinobacteria (10%) and Bacteroidetes (3%). The Cyanobacteria proportion in this lake was very low compared to the proportion found in Salina Verde (1% and 80%, respectively).

Taxonomic Classification bar chart by Genus

Figure 2. Taxonomic Classification bar chart by Genus.

The main cyanobacterial genus in Salina Verde lake was Nodularia (40%), a filamentous nitrogen-fixing group of cyanobacteria commonly associated with algae bloom events in salinic water systems. Other cyanobacterial genus found in Salina Verde lake were Nostoc, Calothrix, Anabaena or Fischerella. None of them was found in significant abundances in Salina Preta.

Figure 3. Nodularia spumigena (source: Biological and Satellite Oceanography Laboratory,

The dominant genera found in Salina Preta were Pseudomonas (7%), Hydrogenophaga (4%), Acidovorax (3%) and Burkholderia (3%).

OmicsBox allows visualizing the taxonomic classification in an interactive manner via Krona charts, in which the different taxonomies are displayed hierarchically. This kind of graphs makes easier to compare samples with each other.

Krona chart comparison between PABKrona chart comparison between VAB

Figure 4. Krona chart comparison between PAB (left) and VAB (right).

4. Functional Characterization

4.1. Assembly and Gene Finding

After comparing the taxonomic compositions, we decided to perform a functional characterization to describe the divergences in metabolism of both microbial communities.

For that purpose, we used MEGAHIT (Metagenomics > Metagenomic Assembly > MEGAHIT) to assemble each sample, and we used FragGeneScan (Metagenomics > Metagenomic Gene Prediction > FragGeneScan) to find genes and gene fragments in these assemblies.

AssemblyN. contigs133631660773998727
Gene FindingPredicted genes71557766404288059839
Avg. gene length (nt)389.79404.66394.01381.19

4.2. Functional Annotation

Genes found after running the gene finding analysis were functionally annotated in three different ways: via PfamScan, via EggNOG Mapper (Metagenomics > Annotation Tools) and via Blast2GO methodology. The Blast2GO annotation was performed by using CloudBlast with BlastP against the NCBI nr (non-redundant) database (Functional Analysis > Blast) using the amino acid sequences from the gene finding step as queries, and the GO mapping (Functional Analysis > Blast2GO Mapping) and the annotation (Functional Analysis > Blast2GO Annotation) steps were executed with default parameters.

As BlastP was performed against the nr database, we decided to check the Top-Hit Species Distribution. When comparing these species to the species found in the initial taxonomic classification, both results were very similar.

In Salina Verde, the Genus Nodularia represents the first four top-hit species. In particular, N. spumigena is the principal species found, which coincides with the Krona chart information. 

In Salina Preta, Acidobacteria or Hydrogenophaga are present in the top-hit species and in the taxonomic classification graphs. 

Top-Hit Species Distribution charts: PAB

Top-Hit Species Distribution charts: VAB

Figure 5. Top-Hit Species Distribution charts: PAB and VAB samples.

Each annotation approach generates different GO terms, so we merged all the information for each sample. After this step, we had four annotations, one per sample. The functional analysis results are summarized here:

Predicted genes71557766404288059839
EggNOGAnn. genes1807519318933314379
% annot.25.26%25.21%21.77%24.03%
PfamScanF. structures31651362851519534736
% charact.44.23%47.34%35.44%58.04%
Blast2GOAnn. genes48061553722729135725
% annot.67.16%72.25%63.65%59.70%
FINALAnn. genes66139746943662850123
% annot.92.43%97.46%85.42%83.76%

Finally, we obtained a sample comparison chart (Metagenomics > Comparative Analysis > Sample Comparison Chart) with the final annotations to examine and compare the functional composition of each lake.

Sample Comparison Chart

Figure 6. Sample Comparison Chart.

In the Salina Verde metagenome, functions related to photosynthesis (and photosynthesis itself) were especially abundant: macromolecule metabolism phosphorous metabolism or pigment biosynthesis. Some Cyanobacteria species had the ability to fix the environmental nitrogen (N. spumigena or Anabaena), and therefore functions like Cellular nitrogen compound metabolism are also abundant.

The above mentioned functions were less abundant in Salina Preta lake. Here, functions related to the use of organic acids as substrates are notably important because of the presence of organisms like Acidovorax. Other functions like response to stress or sulfur compound metabolic process are strongly represented in this lake due to the metabolism of this community.

5. Measured times and Cloud Units consumption

  • Taxonomic classification: 15 min per sample
  • Metagenomic assembly: 5 min per sample
  • Gene prediction: 10 min per sample
  • Functional Annotation
    • PfamScan: 30 min per sample
    • EggNOG-Mapper: 2 hours per sample
    • BlastP against NR: between 1 and 5.5 hours per sample (1.5M Cloud Units in total)

The measured times are provided as maximum times per sample because the analysis can be run in parallel on all 4 samples at the same time. The workflow depicted in figure 7 shows part of the analysis that was run on each sample. Taking advantage of concurrent processing for all 4 samples, we can complete the whole analysis in around 6 hours. Measured times of course depend on samples sizes.

Complete metagenomics workflow

Figure 7. Complete metagenomics workflow.

6. Conclusions

  • Our metagenomic analysis with OmicsBox allowed us to identify the microbial communities living in the lakes and to measure their functional genetic potential.
  • Both taxonomic and functional compositions of Salina Preta and Salina Verde lakes were very different, mainly because of the cyanobacterial blooms which occur in Salina Verde.
  • Some organisms in Salina Verde had the potential to carry out photosynthesis and nitrogen-fixation processes. These functions are not important in Salina Preta lake.

7. References

  • OmicsBox – Bioinformatics made easy. BioBam Bioinformatics. March 3, 2019.
  • Andreote Ana P. D., Dini-Andreote Francisco, Rigonato Janaina, Machineski Gabriela Silva, Souza Bruno C. E., Barbiero Laurent, Rezende-Filho Ary T., Fiore Marli F. (2018). Contrasting the Genetic Patterns of Microbial Communities in Soda Lakes with and without Cyanobacterial Bloom. Frontiers in Microbiology, 9, doi: 10.3389/fmicb.2018.00244.
  • Conesa A., Götz S., Garcia-Gomez JM., Terol J., Talon M. and Robles M. (2005). Blast2GO: a universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics (Oxford, England), 21(18), 3674-6.
  • Conesa A. and Götz S. (2008). Blast2GO: A comprehensive suite for functional analysis in plant genomics. International journal of plant genomics, 2008, 619832.
  • Götz S., Garcia-Gomez JM., Terol J., Williams TD., Nagaraj SH., Nueda MJ., Robles M., Talon M., Dopazo J. and Conesa A. (2008). High-throughput functional annotation and data mining with the Blast2GO suite. Nucleic acids research, 36(10), 3420-35.
  • Huerta-Cepas J., Forslund K., Coelho LP., Szklarczyk D., Jensen LJ., von Mering C. and Bork P. (2017). Fast Genome-Wide Functional Annotation through Orthology Assignment by eggNOG-Mapper. Molecular biology and evolution, 34(8), 2115-2122.
  • Langmead B. and Salzberg SL. (2012). Fast gapped-read alignment with Bowtie 2. Nature methods, 9(4), 357-9.
  • Li D., Liu CM., Luo R., Sadakane K. and Lam TW. (2015). MEGAHIT: an ultra-fast single-node solution for large and complex metagenomics assembly via succinct de Bruijn graph. Bioinformatics (Oxford, England), 31(10), 1674-6.
  • Li D., Luo R., Liu CM., Leung CM., Ting HF., Sadakane K., Yamashita H. and Lam TW. (2016). MEGAHIT v1.0: A fast and scalable metagenome assembler driven by advanced methodologies and community practices. Methods (San Diego, Calif.), 102, 3-11.
  • Mistry J., Bateman A. and Finn RD. (2007). Predicting active site residue annotations in the Pfam database. BMC bioinformatics, 8, 298.
  • Rho M., Tang H. and Ye Y. (2010). FragGeneScan: predicting genes in short and error-prone reads. Nucleic acids research, 38(20), e191.
  • Wood DE. and Salzberg SL. (2014). Kraken: ultrafast metagenomic sequence classification using exact alignments. Genome biology, 15(3), R46.

For further information about how to perform metagenomics in OmicsBox, please consult the user manual.

nodularia (2)
Share on facebook
Share on twitter
Share on linkedin

Blog Categories:


Releases, Media, Announcements, etc.

Use Cases, Reviews, Tutorials

Product Tutorial, Quickstarts, New Features, etc.

Video Tutorials

Helpful Features, Tips and Tricks

Tips And Tricks

Mini-tutorials for common use-cases and to address frequently asked questions FAQs

Most Popular:

Share on facebook
Share on twitter
Share on linkedin
Share on email
Share on print