|
iVar
|
| Command | Description |
|---|---|
| trim | Trim reads in aligned BAM |
| variants | Call variants from aligned BAM file |
| filtervariants | Filter variants across replicates or multiple samples aligned using the same reference |
| consensus | Call consensus from aligned BAM file |
| getmasked | Detect primer mismatches and get primer indices for the amplicon to be masked |
| removereads | Remove reads from trimmed BAM file |
| version | Show version information |
| trimadapter | (EXPERIMENTAL) Trim adapter sequences from reads |
To view detailed usage for each command type ivar <command> Note : Commands maked (EXPERIMENTAL) are still under active development.
iVar uses primer positions supplied in a BED file to soft clip primer sequences from an aligned and sorted BAM file. Following this, the reads are trimmed based on a quality threshold(Default: 20). To do the quality trimming, iVar uses a sliding window approach(Default: 4). The windows slides from the 5' end to the 3' end and if at any point the average base quality in the window falls below the threshold, the remaining read is soft clipped. If after trimming, the length of the read is greater than the minimum length specified(Default: 30), the read is written to the new trimmed BAM file.
Please note that the strand is taken into account while doing the trimming so forward primers are trimmed only from forward strand and reverse primers are trimmed from reverse strand.
To sort and index an aligned BAM file (OPTIONAL, if index is not present iVar will create one), the following command can be used,
Note: All the trimming in iVar is done by soft-clipping reads in an aligned BAM file. This information is lost if reads are extracted in fastq or fasta format from the trimmed BAM file.
Command:
Example Usage:
The command above will produce a trimmed BAM file test.trimmed.bam after trimming the aligned reads in test.bam using the primer positions specified in test_primers.bed and a minimum quality threshold of 15, minimum read length of 50 and a sliding window of 4.
Example Usage:
The command above will allow you to go from alignment to consensus sequence in a single command using the bwa aligner.
Example BED file
iVar uses the output of the samtools mpileup command to call variants - single nucleotide variants(SNVs) and indels. In order to call variants correctly, the reference file used for alignment must be passed to iVar using the -r flag. The output of samtools pileup is piped into ivar variants to generate a .tsv file with the variants. There are two parameters that can be set for variant calling using iVar - minimum quality(Default: 20) and minimum frequency(Default: 0.03). Minimum quality is the minimum quality for a base to be counted towards the ungapped depth to canculate iSNV frequency at a given position. For insertions, the quality metric is discarded and the mpileup depth is used directly. Minimum frequency is the minimum frequency required for a SNV or indel to be reported.
iVar can identify codons and translate variants into amino acids using a GFF file in the GFF3 format containing the required coding regions (CDS). In absence of a GFF file, iVar will not perform the translation and "NA" will be added to the output file in place of the reference and alternate codons and amino acids. The GFF file in the GFF3 format can be downloaded via ftp from NCBI RefSeq/Genbank. They are usually the files with the extension ".gff.gz". For example, the GFF file for Zaire Ebolavirus can be found here. More details on GFF3 files hosted by NCBI can be found in their ftp FAQs.
Some RNA viruses such as Ebola virus, might have polymerase slippage causing the insertion of a couple of nucleotides. More details can be found here. iVar can account for this editing and identify the correct open reading frames. The user will have to specify two additional parameters, EditPosition: Position at which edit occurs and EditSequence: The sequence tht is inserted at the given positon, in the "attributes" column of the GFF file to account for this. A test example is given below,
If a certain base is present in multiple CDSs, iVar will add a new row for each CDS frame and distinguish the rows by adding the ID (specified in attributes of GFF) of the GFF feature used for the translation. This is shown for position 42 in the example output below. There are two rows with two different GFF features: id-test3 and id-test4.
Command:
Example Usage:
The command above will generate a test.tsv file.
Example of output .tsv file.
Description
| Field | Description |
|---|---|
| REGION | Region from BAM file |
| POS | Position on reference sequence |
| REF | Reference base |
| ALT | Alternate Base |
| REF_DP | Ungapped depth of reference base |
| REF_RV | Ungapped depth of reference base on reverse reads |
| REF_QUAL | Mean quality of reference base |
| ALT_DP | Ungapped depth of alternate base. |
| ALT_RV | Ungapped deapth of alternate base on reverse reads |
| ALT_QUAL | Mean quality of alternate base |
| ALT_FREQ | Frequency of alternate base |
| TOTAL_DP | Total depth at position |
| PVAL | p-value of fisher's exact test |
| PASS | Result of p-value <= 0.05 |
| GFF_FEATURE | ID of the GFF feature used for the translation |
| REF_CODON | Codong using the reference base |
| REF_AA | Amino acid translated from reference codon |
| ALT_CODON | Codon using the alternate base |
| ALT_AA | Amino acid translated from the alternate codon |
Note: Please use the -B options with samtools mpileup to call variants and generate consensus. When a reference sequence is supplied, the quality of the reference base is reduced to 0 (ASCII: !) in the mpileup output. Disabling BAQ with -B seems to fix this. This was tested in samtools 1.7 and 1.8.
iVar can be used to get an intersection of variants(in .tsv files) called from any number of replicates or from different samples using the same reference sequence. This intersection will filter out any iSNVs that do not occur in a minimum fraction of the files supplied. This parameter can be changed using the -t flag which range from 0 to 1 (default). Fields that are different across replicates(fields apart from REGION, POS, REF, ALT, REF_CODON, REF_AA, ALT_CODON, ALT_AA) will have the filename added as a suffix. If there are a large number of files to be filtered, the -f flag can be used to supply a text file with one sample/replicate variant .tsv file per line.
Command:
Example Usage: The command below only retains those variants that are found in atleast 50% of the fiels supplied
The three replicates can also be supplied using a text file as shown below
filter_files.txt
The command above will prodoce an output .tsv file test.filtered.tsv.
Example output of filtered .tsv file from three files test_rep1.tsv and test_rep2.tsv
Description of fields
| No | Field | Description |
|---|---|---|
| 1 | REGION | Common region across all replicate variant tsv files |
| 2 | POS | Common position across all variant tsv files |
| 3 | REF | Common reference base across all variant tsv files |
| 4 | ALT | Common alternate base across all variant tsv files |
| 5 | GFF_FEATURE | GFF feature used for the translation |
| 6 | REF_CODON | The codon using the reference base |
| 7 | REF_AA | Reference codon translated into amino acid |
| 8 | ALT_CODON | Codon using the alternate base |
| 9 | ALT_AA | Alternate codon translated into amino acid |
| 10 | REF_DP_<rep1-tsv-file-name> | Depth of reference base in replicate 1 |
| 11 | REF_RV_<rep1-tsv-file-name> | Depth of reference base on reverse reads in replicate 1 |
| 12 | REF_QUAL_<rep1-tsv-file-name> | Mean quality of reference base in replicate 1 |
| 13 | ALT_DP_<rep1-tsv-file-name> | Depth of alternate base in replicate 1 |
| 14 | ALT_RV_<rep1-tsv-file-name> | Deapth of alternate base on reverse reads in replicate 1 |
| 15 | ALT_QUAL_<rep1-tsv-file-name> | Mean quality of alternate base in replicate 1 |
| 16 | ALT_FREQ_<rep1-tsv-file-name> | Frequency of alternate base in replicate 1 |
| 17 | TOTAL_DP_<rep1-tsv-file-name> | Total depth at position in replicate 1 |
| 18 | PVAL_<rep1-tsv-file-name> | p-value of fisher's exact test in replicate 1 |
| 19 | PASS_<rep1-tsv-file-name> | Result of p-value <= 0.05 in replicate 1 |
| 20 | Continue rows 10 - 19 for every replicate provided |
To generate a consensus sequence iVar uses the output of samtools mpileup command. The mpileup output must be piped into ivar consensus. There are five parameters that can be set - minimum quality(Default: 20), minimum frequency threshold(Default: 0), minimum depth to call a consensus(Default: 10), a flag to exclude nucleotides from regions with depth less than the minimum depth and a character to call in regions with coverage lower than the speicifed minimum depth(Default: 'N'). Minimum quality is the minimum quality of a base to be considered in calculations of variant frequencies at a given position. Minimum frequency threshold is the minimum frequency that a base must match to be called as the consensus base at a position. If one base is not enough to match a given frequency, then an ambigious nucleotide is called at that position. Minimum depth is the minimum required depth to call a consensus. If '-k' flag is set then these regions are not included in the consensus sequence. If '-k' is not set then by default, a 'N' is called in these regions. You can also specfy which character you want to add to the consensus to cover regions with depth less than the minimum depth. This can be done using -n option. It takes one of two values: '-' or 'N'.
As an example, consider a position with 6As, 3Ts and 1C. The table below shows the consensus nucleotide called at different frequencies.
| Minimum frequency threshold | Consensus |
|---|---|
| 0 | A |
| 0.5 | A |
| 0.6 | A |
| 0.7 | W(A or T) |
| 0.9 | W (A or T) |
| 1 | H (A or T or C) |
If there are two nucleotides at the same frequency, both nucleotides are used to call an ambigious base as the consensus. As an example, consider a position wiht 6 Ts, 2As and 2 Gs. The table below shows the consensus nucleotide called at different frequencies.
| Minimum frequency threshold | Consensus |
|---|---|
| 0 | T |
| 0.5 | T |
| 0.6 | T |
| 0.7 | D(A or T or G) |
| 0.9 | D(A or T or G) |
| 1 | D(A or T or G) |
The output of the command is a fasta file with the consensus sequence and a .txt file with the average quality of every base used to generate the consensus at each position. For insertions, the quality is set to be the minimum quality threshold since mpileup doesn't give the quality of bases in insertions.
Command:
Example Usage:
The command above will produce a test.fa fasta file with the consensus sequence and a test.qual.txt with the average quality of each base in the consensus sequence.
iVar uses a .tsv file with variants to get the zero based indices(based on the BED file) of mismatched primers. This command requires another .tsv file with each line containing the left and right primer names separated by a tab. This is used to get both the primers for an amplicon with a single mismatched primer. The output is a text file with the zero based primer indices delimited by a space. The output is written to a a text file using the prefix provided.
Command:
Example BED file
Example primer pair information file
Example Usage:
The command above produces an output file - test.masked.txt.
Example Output:
This command accepts an aligned and sorted BAM file trimmed using ivar trim and removes the reads corresponding to the supplied primer indices, which is the output of ivar getmasked command. Under the hood, ivar trim adds the zero based primer index(based on the BED file) to the BAM auxillary data for every read. Hence, ivar removereads will only work on BAM files that have been trimmed using ivar trim.
Command:
Example Usage:
The ivar trim command above trims test.bam and produced test.trimmed.bam with the primer indice data added. The ivar removereads command produces an output file - test.trimmed.masked.bam after removing all the reads corresponding to primer indices - 1,2,7 and 8.
Note: This feature is under active development and not completely validated yet.
trimadapter in iVar can be used to trim adapter sequences from fastq files using a supplied fasta file.
1.9.5