Analysis
The objective of the analysis is to replicate the results and to add new results to the data. Assembly is first step to understand and predict the mechanisms involved in an organism. Genome assembly is carried on the PacBio reads from the article Christel et al. This is the starting point of both this project and the article Christel et al. The assembled genome is then further used for predicting genes present and to functionally annotate the same.
The software used for assembly is Canu, which supports assembly of PacBio read. The software does trimming of the data and correction on it. The trimmed reads of more than 1000 bp length and 0.045 fraction of overlap is used for assembly. The software performs quality assessment of the reads, therefore the raw reads are given as input without any quality assessment.
The command used for canu is:
canu -p LeptospirillumT -d LT-mix genomeSize=2.5m stopOnReadQuality=false maxThreads=4 -pacbio-raw /home/akshaips/genome_analysis/raw_data/DNA/ERR2028???.fastq.gz
-p to set the file name prefix of intermediate and output files
-d to set the directory name for the output files
genomeSize is obtained from the paper and given as 2.5 million bases
stopOnReadQuality Canu evaluates the quality of the reads to be too low and stops the assembly, to avoid this it is set to FALSE
maxThreads determines the number of threads used by the software for calculation
There were 215276 raw reads with 2998404754 bases having a coverage of 1199.36x. After the quality check and trimming, there were 10774 reads with 132551771 bases having a coverage of 53.02x. This shows that 95% of the data is omitted due to low quality, this my be due to experimental issues with sequencing and this was compensated by the large dataset which can cover the required genome size.
The final result contains 7 contigs, with largest one having 2567031 bases long. This is little shorter than the sequence from the article, which has 2569357 bases. The quality of the assembly is analysed in the later stages.
The quality of the assembly is checked using Quast. This is important to do before going ahead with further analysis. Quast can be run with or without the reference genome, but works best with reference genome. The reference genome for our analysis is obtained from the article.
The command used for quast is:
quast.py LeptospirillumT.contigs.fasta -R OBMB01.fasta -t 4
-R takes the reference genome
-t the number of threads used for the analysis
The result shows that 2 contigs aligned with the reference genome but in two pieces and hence it was classified as misassembled. The rest 5 contigs are either did not align with anything and therefore would not be used for further analysis.
The largest contig is aligned with reference genome and from the dot plot it shows that there are repeating regions. There is one in the starting of the genome, this might be due to overlapping of the circular genome. The other one is inverted one at the end of the sequence.
Figure 1. Dotplot showing alignment between the 2 contigs from Quast and the genome from the article by Christel et al.
From the 2 contigs, the highlighed region shows the contig of length 42097, which is overlapping with the largest contig itself. Hence the smaller contig was removed from further analysis.
The larger contig is extracted script and used for to generate dotplot.
Figure 2. Dotplot showing alignment between the largest contig and reference genome from the article. The repeated region in the beginning and inverted repeat in the beginning of the sequence can seen.
To see the repeating region at the beginning of the contig, dotplot is generated for the largest contig with itself. It is seen that around ~20000 bp is repeating in the contig.
Figure 3. Dotplot of largest contig with itself, which shows the repeating region at the top right and bottom left corner (which are start and end of the sequence).
It is important to remove this region. To find the location of repeats on the genes, NUCmer from MUMmer was used. The result of the NUCmer is used to identify the repeating region.This was later removed by samtools.
The command used for nucmer and samtools are:
nucmer -maxmatch -nosimplify LeptospirillumT.contigs.fasta LeptospirillumT.contigs.fasta
show-coords -lrcTH out.delta > overlapping_coords.out
samtools faidx largest_contig.fasta tig00000001:1-2546870 > largest_contig_after_trimming.fasta
| Start 1 | End 1 | Start 2 | End 2 | Length of match 1 | Length of match 2 | Percentage match | Contig number | Contig number |
|---|---|---|---|---|---|---|---|---|
| 1 | 20164 | 2546870 | 2567031 | 20164 | 20162 | 99.89 | tig00000001 | tig00000001 |
| 2546870 | 2567031 | 1 | 20164 | 20162 | 20164 | 99.89 | tig00000001 | tig00000001 |
| 1130099 | 1135576 | 1205517 | 1200041 | 5478 | 5477 | 99.93 | tig00000001 | tig00000001 |
| 1200041 | 1205517 | 1135576 | 1130099 | 5477 | 5478 | 99.93 | tig00000001 | tig00000001 |
Table 1. Table showing important data from the analysis. It can be seen in the first entry, the starting region is repeated in the end. There are other repeats in the middle which are not removed, they might be random repeats or functionally important part of the genome. The complete table can be found here
The output of the samtools is used to create dotplot to see the result.
Figure 4. Dotplot of largest contig after trimming with itself. It can be seen that the repeating region has been removed.
The result plot does not show repeats at the beginning or end of the sequence. Even though there are repeats in the middle of the sequence, they were kept as it is. The contig length decreased to 2546870 which is 22487 bp less than the largest contig from the paper. The output fasta is further used for annotation of genes.
Annotation is the step to find the genes and to predict its function. It is one of the crucial steps, in order to understand the genes present in the organism, which will in turn help to predict the features of the organism such as ability to metabolize iron and other metals, acidophilic nature etc. Annotation locates the gene in the genome and this can be used to find coding region and its transcription product. The software used for annotation is Prokka, which can be then validated using Eggnoger.
The command used for prokka is:
prokka --prefix lferr --addgenes --genus Leptospirillum --species ferriphilum --strain DSM_14647 --gram neg --usegenus --proteins Leptospirillum --rfam largest_contig_after_trimming.fasta --force --cpus 0
_--prefix gives prefix names for the output files
--addgenes to identify gene name for the CDS
--genus input genus name for prokka
--species input species name for prokka
--strain input strain name for prokka
--gram input for prokka whether the organism is gram-positive or gram-negative
--usegenus input to specify to use genus specific database
--proteins Fasta file of trusted proteins to first annotate from
--rfam tells Prokka to also search for ncRNA's+
--force deletes the output file, if already present before the analysis
--cpus number of threads to use, 0 used the max amount available
Pokka was ran and it identified 2586 genes of which 2523 codes for protein. The result is almost similar to the results from the article, where they have 2541 genes with 2,486 coding sequence. The difference in numbers is due prokka parameters used to identify these genes.
The results from prokka is validated using EggNOG-mapper. In this project we have used the online server on the output data given by prokka. The output of prokka containing all the proteins in nucleotide format is used as input for EggNOG-mapper.
The results of prokka and EggNOG-mapper matches in majority of cases and also for some proteins, EggNOG-mapper did not give any results. These results are to be compared with article data. From the dataset given by the article, lf_chrm.gbk file contains all the CDS (2541 CDS) and also their prokka annotation. These information is extracted from the file using this script. The sequences used for blast are amino acid sequences.
For to map our results and article results, BLAST+ is used. The standalone version is used in our case to perform bi-directional blast to map two genomes.
The steps for bidirectional blast includes,
- Making database to search, the genes from article and genes from our analysis are made into two blast databases using this command. (These are run in two separate directories)
makeblastdb -in lferr.faa -input_type fasta -dbtype prot -hash_index # this is for our genes from prokka output
makeblastdb -in genes_from_article.fasta -input_type fasta -dbtype prot -hash_index #this is the output of above python code
-in is the input fasta file
-input_type file format
-dbtype type of database, the sequences are amino acid sequences hence prot
-hash_index to index the output database
- Our genes are searched against database of genes from article and vice versa. The commands used are.
blastp -db my_genes_db/lferr.faa -num_threads 8 -max_target_seqs 1 -query article_genes.fasta -out article_genes_against_my_genes.txt -evalue 0.001 -outfmt " 6 qseqid sseqid pident length qlen slen mismatch gapopen qstart qend sstart send evalue bitscore"
blastp -db article_genes_db/article_genes.fasta -num_threads 8 -max_target_seqs 1 -query article_genes.fasta -out article_genes_against_my_genes.txt -evalue 0.001 -outfmt " 6 qseqid sseqid pident length qlen slen mismatch gapopen qstart qend sstart send evalue bitscore"
The output is in table format and only the best hits (-max_target_seqs 1) are chosen.
-num_threads is for specifying number of threads to be used
-db is the database to search against
-out is the output file name
-query is the input query sequences
-outfmt is the type of output file. 6 specifies table with the parameters specified in quotes as output
-max_target_seqs number of output needed for each query sequences
-evalue specify the e-evalue
- The output file from both blast output is mapped using this script. The results from the scripts contain all the sequences having bidirectional hits, single way hits and which did not perform blast. The script is ran using
python3 check_validity_of_blast.py my_genes_against_article.txt article_genes_against_my_genes.txt my_genes.faa article_genes.fasta egg_nog_job_MM_wzm4hbmt_annotations.tsv
The result shows that there are 2434 bidirectional blast hits, 75 one way hits and 14 of them did not find any blast results. The 14 genes are additional genes which are found in our analysis but not in the article. The function of the are not known (hypothetical protein). In the 75 hits there are 42 hypothetical protein and from 2434 proteins there are 1028 hypothetical proteins by prokka annotation.
Comparing with the article results and EggNOG results, some of the hypothetical proteins have been annotated. In most of the cases the EggNOG results and prokka results from the article are matching. From the results it is interesting to note the important characteristics of the organism such as, metal resistance genes and metal transportation genes. There are multiple metal resistance genes and efflux pumps are present in the organism. Some results are.
| my gene ID | prokka annotation | article gene ID | article prokka annotation | eggnog annotation |
|---|---|---|---|---|
| CBNIABCL_02562 | hypothetical protein | LFTS_00132 | uracil-DNA glycosylase, family 4 | Uracil DNA glycosylase superfamily |
| CBNIABCL_00625 | Cobalt-zinc-cadmium resistance protein CzcA | LFTS_02037 | heavy metal efflux pump, CzcA family | AcrB/AcrD/AcrF family |
| CBNIABCL_02373 | Copper-exporting P-type ATPase A | LFTS_00313 | Cu+-exporting ATPase | E1-E2 ATPase |
| CBNIABCL_00699 | hypothetical protein | LFTS_01966 | Cytochrome C and Quinol oxidase polypeptide I | heme-copper terminal oxidase activity |
Table 2. Annotation of genes from the new analysis and from the article
(Complete data from EGGNoG contains KEGG pathway data, Gene Ontology IDs etc. Which is useful during RNA mapping and is present here)
Full results of the annotation is present in here. This will help in mapping differential expression data with the genome.
In this section, synteny of genes were analysed between L. ferriphilum and L. ferrooxidans. From this it is possible to understand how the organisms have evolved differently. Full genome of both the organisms are used for the analysis. Satsuma2 was used to compare the synteny, from which it was visualized using MizBee. Then synteny was then generated using Artemis. L. ferriphilum genome is blasted against L. ferrooxidans, for generating the data input for Artemis.
Figure 5. At first,Mizbee graph for synteny between L. ferriphilum and L. ferrooxidans. Second is Synteny graph using Aretemis.
From these images, it is easy to see the shuffling of genes between the two species. It is clear that the organisms have undergone good chunk of changes since they have split into two species. The blue lines in the artemis shows inversion of genes in the species.
Figure 6. At first,Mizbee graph for synteny between L. ferriphilum and L. ferriphilum ML04 strain. Second is Synteny graph using Aretemis.
The analysis shows that the genes are matching with the reverse strand of the L. ferriphilum ML04. Which shows that the synteny is seen between these two strains, but there are some genes that are not matching.
As we analysed genome for identifying genes, these are not enough to understand how the organism survive on extreme conditions. To analyse how the organism uses its system for growth, Christel et al conducted growth experiments on Fe2+ containing continous culture and on culture grown on charcopyrite (CuFeS2). A differential expression analysis can be used to understand the role of genes and proteins in each condition for the growth of the organism. The RNA reads are available in the paper which are used for our analysis.
The first step of analysis is to remove illumina adapters from the reads, this can be done using trimmomatic. Since trimmed reads are provided in the paper, further analysis are carried out on these reads.
After trimming generally, the quality of trimming has to be assessed. Here, we can verify the quality of trimming by comparing raw-reeds and trimmed reads. FastQC is used for quality assessment.
The results are kept on different folders to seggregate easily.
The commands used are:
fastqc -o fastqc_output -f fastq -t 4 /domus/h1/akshaips/genome_analysis/raw_data/RNA/*
fastqc -o fastqc_output -f fastq -t 4 /domus/h1/akshaips/genome_analysis/raw_data/RNA_trimmed_reads/*
The output of FastQC contains result in HTML format. To collate the results, MultiQC is used. MultiQC combines the plot to a single plot, which are shown below.
Figure 7. Graph showing the phred quality score for the raw reads for each position. The red line indicates bad scores and green for good scores. Many reads had good scores.
Figure 8. Graph showing the phred quality score for the trimmed reads for each position. The quality seem to be improved and no results fall below score of 28.
Figure 9. Graph showing the adapter content of the raw reads. It can be seen that, there is considerable amount of adapter in the reads.
Figure 10. Graph showing the adapter content of the trimmed reads. It shows that most of the adapter has been removed and a few sequences are present, which can be neglected.
It can be seen that trimming improved the quality of reads and removed most of the adapter content. Hence trimmed reads can be used for further analysis.
Using trimmed reads, we can align/map to our genome and count the number of alignment on each gene. We can use Burrows Wheeler Alignment for this purpose, using BWA tool. The algorithm we can use is BWA-MEM, which is for sequences between 70bp to 1 Mbp in length.
The output of BWA produces alignement in SAM or BAM files, which can be sorted using samtools. This sorted alignment then can be used in HTSeq to get the count information of each gene. HTSeq requires GFF file, which is one of the output file of prokka (Remove all the fasta sequences in the GFF file). To produce alignment using BWA, we have to index our genome. The genome used here is output fna given by prokka.
The command for indexing is.
bwa index -p index_file.fasta lferr.fna
-p the name of the output file, has to end with fasta
From the differential expression experiment we have 10 results. Only 5 of them are used since the other five were replicates (later it was found that they were multiple copies of the same file). The experimental files were mapped with SRA id from ncbi using table 3 in the paper.
| Experiment ID | File ID | Culture type |
|---|---|---|
| LNU-LXX9-Si00-CnA-P-B6-R1 | ERR2036633 | Continuous |
| LNU-LXX-Si00-14C-P-R1 | ERR2036632 | Batch,mineral |
| LNU-LXX-Si00-14B-P-R1 | ERR2036631 | Batch,mineral |
| LNU-LXX9-Si00-CnA-P-B1-R1 | ERR2036630 | Continuous |
| LNU-LXX9-Si00-CnA-P-B7-R1 | ERR2036629 | Continuous |
| LNU-LXX9-Si00-CnA-P-B6-R1 | ERR2117292 | Continuous |
| LNU-LXX-Si00-14C-P-R1 | ERR2117291 | Batch,mineral |
| LNU-LXX-Si00-14B-P-R1 | ERR2117290 | Batch,mineral |
| LNU-LXX9-Si00-CnA-P-B1-R1 | ERR2117289 | Continuous |
| LNU-LXX9-Si00-CnA-P-B7-R1 | ERR2117288 | Continuous |
Table 3. Data from the paper and corresponding information
For each of the growth condition (5 Experiments), BWA is used. Since the output files are huge, these are piped to other tools directly.
The command for BWA is. (The command is looped for all the five variants)
bwa mem -t 8 ../index_file.fasta /domus/h1/akshaips/genome_analysis/raw_data/RNA_trimmed_reads/ERR2036629_P1.trim.fastq.gz /domus/h1/akshaips/genome_analysis/raw_data/RNA_trimmed_reads/ERR2036629_P2.trim.fastq.gz | samtools sort | htseq-count -t CDS -i ID -f bam -s no -r pos - lferr_prokka_gff.gff > output_2036629.txt
-t on the bwa is for number of threads
-t on htseq is to count which has CDS feature
-i format of the identity identifier ##
-s no tells it to consider reads that map on the either strand, not just the correct one
-r pos tells HTSeq that input file is sorted BAM file
-f format of the input file
The analysis is only conducted for paired reads output of trimmomatic, since they contain most of the reads and also would contain high quality reads. The output result contains 10 files (since ran of 10, run only on 5 since they are file duplicates). Each result contains count of the genes expressed and also details about ambiguous match, low quality sequences, reads which are not aligned etc.
With the above results, we can conduct differential expression analysis.
In this section, differential expression analysis is conducted to find the difference in gene expression between the two conditions (continuous culture with iron and bioleaching cultures containing chalcopyrite(CuFeS2)) in which the organism have grown. For this analysis, output of htseq is used in a R package called DESeq2. The input for DESeq2 is created as from the manual. The two input files for the DESeq2 are table with conditon and count table. The R script used for DESeq2 is here.
out of 2518 with nonzero total read count
adjusted p-value < 0.1
LFC > 0 (up) : 386, 15%
LFC < 0 (down) : 381, 15%
outliers [1] : 0, 0%
low counts [2] : 98, 3.9%
(mean count < 9)
[1] see 'cooksCutoff' argument of ?results
[2] see 'independentFiltering' argument of ?results
It is seen that arount ~760 genes are differentially expressed, which almost 1/3 of the total genes present in the organism. A PCA analysis is conducted to find the difference between them.
Figure 11. Data of PCA analysis from differential expression analysis on the two cultures.
From the plot it is seen that 86% of the variance is covered by PC1, which tells about the difference between the two types of cultures. This supports differential expression of the organism in different conditions. The PC2 having 9% of the variation explain about, how different they are between the samples. This suggests that, majority of the difference is between the cultures and there is a small fraction of difference between samples due to experimental conditions, environmental conditions etc.
In this study conducted on L. ferriphilum, the aim was to replicate and compare the results from the paper by Christel et al. From the study it was shown that majority of the results from the paper matches our study. The difference in the analysis and parameters used for analysis would count for changes in the results. The major change that can occur is during assembly of the genome and predicting the genes from it. It was seen that most of the genes from the paper can be seen in our analysis, but there are some genes which are exclusive to each analysis. From later analysis it can be seen that the major role playing genes and their expression is almost identical in both our study and from the paper. KEGG annotation shows presence of all the genes annotated from the paper and their importance in the organism. This gave better insight about how the organism is able to survive on extreme conditions such as metal concentrated, low Ph etc. In conclusion, the study gave us almost identical results, which validates both our method and studies and also the data from the paper.
The genes present in out organism belongs to certain pathways, which can be analysed using KEGG mapper. The output of Eggnog contains the KEGG ID for the genes. These are extracted using this script. The output of the script is used as the input for the KEGG mapper.
This analysis is done to understand the capabilities of L. ferriphilum. It was shown that they survive in extreme pH and can survive in metal rich environment. L. ferriphilum also forms biofilm, which is a crucial part in biomining. To analyse these features KEGG can be used to map our genes to established pathways.
The below results shows, presence of metal resistant and transporter genes. The resistance to metal is attained by metal efflux pumps in the organism, it can be seen that presence of heavy metal efflux systems for iron,copper,gold,cobalt etc makes the organism suitable for bio-mining activities. Efflux systems coupled with transporters helps in maintaining the concentration of metal content in the organism, which helps them to survive on these metals.
K12346 SMF; metal iron transporter
K03322 mntH; manganese transport protein
K15727 czcB, cusB, cnrB; membrane fusion protein, heavy metal efflux system
K07798 cusB, silB; membrane fusion protein, copper/silver efflux system
K08168 tetB; MFS transporter, DHA2 family, metal-tetracycline-proton antiporter
K15725 czcC, cusC, cnrC; outer membrane protein, heavy metal efflux system
K07796 cusC, silC; outer membrane protein, copper/silver efflux system
K02014 TC.FEV.OM; iron complex outermembrane recepter protein
K19594 gesB, mexQ; gold/copper resistance efflux pump
K11326 nrsA, czcA; cation efflux system protein involved in nickel and cobalt tolerance
K02006 Cobalt/Nickel transporter
K02006 Nickel transporter
It was shown that these organisms can survive at extreme pH of 1.4-1.8. This is facilitated by proton pumps which pumps against the gradient to maintain the concentration within the organism. To maintain pH of 7, H+ antiporters pumps proton against the concentration, the analysis shows presence of H+ antiporter genes in the organism.
K08162 Drug:H+ antiporter-1 (12 spanner) (DHA1) family
K03446 Drug:H+ antiporter-2 (14 spanner) (DHA2) family
Biofilm formation is another important feature of L. ferriphilum, this helps to extract metals from their ores (In this case iron from chalcopyrite (CuFeS2)). The genes correspond to pathways in three organisms, which further proves the ability to form biofilm.
Figure 12. Possible biofilm pathway from Vibrio cholerae
Figure 13. Possible biofilm pathway from Pseudomonas aeruginosa
Figure 14. Possible biofilm pathway from Escherichia coli
L. ferriphilum contain genes for nitrogen fixation. It converts extracellular nitrite to ammonia, which is used up for the production of amino acids along with TCA cycle. This proves that how they survive on iron oxidation as energy source. This can be seen in the almost pathway.
Figure 15. Possible nitrogen fixation pathway in L. ferriphilum.
It was discussed that L. ferriphilum colonize metal sulfide surfaces more efficiently than do the Acidithio- bacilli. This is made possible by motility and chemotaxis. This help them form biofilm on the metal ores, which will enhance the biomining activity. This was seen in the pathway.
Figure 16. Bacterial chemotaxis pathway
Figure 17. Pathway showing flageller assembly
From these analysis, it can be seen that the annotation in the paper matches with our genes. This is not only giving the verification of the results but also deeper understanding of the mechanism inside L. ferriphilum.
In this section, the genes of L. ferriphilum and L. ferrooxidans are compared to find the common genes in them. This would give us an idea about what these organisms have in common. The genome of L. ferrooxidans is obtained from the accession number AP012342. Prokka is used to predict the genes present in them. Using the prokka annotation of L. ferriphilum and L. ferrooxidans, a bidirectional blast is ran using the same steps in the above analysis (Extra analysis : Annotation validation, the scripts used are the same scripts). The eggnog annotation of L. ferriphilum is also used for validity of the analysis. Using the python script mentioned in the above analysis, the genes are compared.
The results gave 1085 bi-directional hits, 291 single way hits and 427 genes having no hits on the L. ferrooxidans genes. There are total of 1084 hypothetical proteins in L. ferriphilum and 878 in L. ferrooxidans. From the 1085 bi-directional hits, hypothetical proteins are removed. Which gave us 1034 genes with annotation. On a top level analysis, most of the genes have same annotated function in L. ferriphilum and L. ferrooxidans (There were very few with different functions, annotated by prokka).
| L. ferriphilum gene | prokka annotation | L. ferrooxidans gene | prokka annotation | L. ferriphilum eggnog annotation |
|---|---|---|---|---|
| CBNIABCL_00110 | Chromosomal replication initiator protein DnaA | DDCPAMKM_00001 | Chromosomal replication initiator protein DnaA | it binds specifically double-stranded DNA at a 9 bp consensus (dnaA box) 5'-TTATC CA A CA A-3'. DnaA binds to ATP and to acidic phospholipids |
| CBNIABCL_00051 | Phosphoglycolate phosphatase | DDCPAMKM_00059 | Phosphoglycolate phosphatase | HAD-hyrolase-like |
| CBNIABCL_02113 | Mercuric reductase | DDCPAMKM_00161 | Mercuric reductase | cell redox homeostasis |
| CBNIABCL_02537 | DNA polymerase III subunit tau | DDCPAMKM_00227 | Holliday junction ATP-dependent DNA helicase RuvB | DNA polymerase III is a complex, multichain enzyme responsible for most of the replicative synthesis in bacteria. This DNA polymerase also exhibits 3' to 5' exonuclease activity |
| CBNIABCL_02007 | Ferrous-iron efflux pump FieF | DDCPAMKM_01749 | Manganese efflux system protein MneS | Dimerisation domain of Zinc Transporter |
| CBNIABCL_00217 | Cation efflux system protein CusB | DDCPAMKM_00345 | Cobalt-zinc-cadmium resistance protein CzcB | Belongs to the membrane fusion protein (MFP) (TC 8.A.1) family |
| CBNIABCL_02522 | Cobalt-zinc-cadmium resistance protein CzcA | DDCPAMKM_00241 | Cobalt-zinc-cadmium resistance protein CzcA | AcrB/AcrD/AcrF family |
| CBNIABCL_00216 | Cobalt-zinc-cadmium resistance protein CzcA | DDCPAMKM_00346 | Cobalt-zinc-cadmium resistance protein CzcA | transmembrane transporter activity |
| CBNIABCL_02359 | Nickel and cobalt resistance protein CnrB | DDCPAMKM_00428 | Multidrug resistance protein MdtA | Biotin-lipoyl like |
| CBNIABCL_02287 | Na(+)/H(+) antiporter NhaH | DDCPAMKM_00098 | K(+)/H(+) antiporter NhaP | Sodium/hydrogen exchanger family |
| CBNIABCL_01531 | Nitrogenase iron protein | DDCPAMKM_02427 | Nitrogenase iron protein | "The key enzymatic reactions in nitrogen fixation are catalyzed by the nitrogenase complex, which has 2 components the iron protein and the molybdenum-iron protein" |
Table 4. Prokka annotation of L. ferriphilum and L. ferrooxidans and eggnog annotation of L. ferrooxidans
From the above table, it can be seen that both the organisms have metal resistance genes and metal efflux pumps which helps them to withstand metal environment. With the help of genes for nitrogen fixation, they can produce energy and amino acids in such environment. Interestingly we can see Na(+)/H(+) antiporter in both the organisms, which depict that both the organisms can live in low pH conditions. These are the quick interesting factors in the comparison study. The compiled output file of blast analysis is present here. The output used for the above analysis after filtering is here.
In this analysis, protein sequences of L. ferriphilum is used for annotation. The prokka output of the L. ferriphilum genome is used as input for Blast2GO. These sequences are blasted against uniref90 database locally, the information of 20 hits are kept for annotation. From the 2524 genes, only 1801 had hits against this protein database. These sequences are then mapped against GO database, for which 1330 sequences had results. The 2524 proteins were annotated using interpro scan, which annotates using protein families, domains and functional sites. The results of mapping and interpro are combined to one annotation to increase the confidence of the prediction. The full data obtained is here.
| SeqName | Description | InterPro GO Names | Prokka annotation |
|---|---|---|---|
| CBNIABCL_00233 | 53612UniRef90_P13510 Cobalt-zinc-cadmium resistance protein CzcB n=8 Tax=Betaproteobacteria RepID=CZCB_CUPMC | C:membrane; F:transmembrane transporter activity; P:transmembrane transport | Cobalt-zinc-cadmium resistance protein |
| CBNIABCL_02522 | 53611UniRef90_P13511 Cobalt-zinc-cadmium resistance protein CzcA n=12 Tax=Bacteria RepID=CZCA_CUPMC | P:cation transport; F:cation transmembrane transporter activity; C:membrane; C:integral component of membrane; F:transmembrane transporter activity; P:transmembrane transport | Cobalt-zinc-cadmium resistance protein CzcA |
| CBNIABCL_02287 | 763192UniRef90_A0A075WDG9 Na+/H+ antiporter, bacterial form n=2 Tax=Archaeoglobus fulgidus RepID=A0A075WDG9_ARCFL | P:cation transport; P:sodium ion transport; P:regulation of pH; F:solute:proton antiporter activity; F:sodium:proton antiporter activity; C:integral component of membrane; P:transmembrane transport | Na(+)/H(+) antiporter NhaH |
| CBNIABCL_02007 | 635780UniRef90_D5E6U7 Cation diffusion facilitator family transporter n=2 Tax=Methanohalophilus mahii RepID=D5E6U7_METMS | P:cation transport; F:cation transmembrane transporter activity; C:integral component of membrane; P:transmembrane transport | Ferrous-iron efflux pump FieF |
| CBNIABCL_01531 | 160352UniRef90_P06117 Nitrogenase iron protein n=916 Tax=root RepID=NIFH_BRADU | F:ATP binding; P:nitrogen fixation; F:nitrogenase activity; F:oxidoreductase activity; P:oxidation-reduction process | The key enzymatic reactions in nitrogen fixation are catalyzed by the nitrogenase complex, which has 2 components the iron protein and the molybdenum-iron protein |
Table 5. Concise Blast2GO results and prokka results. The full results of Gene Ontology can be found in the files attached above.
From the above results it can be seen that most of the protein annotation matches with prokka annotation. This validates the results of the annotation made for the genes using eggnog. It can be seen that H+ antiporters, metal efflux pumps and metal resistance genes are annotated by both prokka/eggnog and blast2go(interpro and GO) results. This validates the presence of these genes in the organisms.