Step 1: consensusProfile

Once the alignment is imported, a consensus profile can be generated. The consensus profile contains all the information needed for the subsequent design process:

• Proportion of the letters A, C, G, T, and other
• Proportion of gaps (-)
• Majority consensus base
• Identity
• IUPAC consensus character
• Coverage

The majority consensus base is the most frequently occurring letter among the letters A, C, G, T and -.

Identity represents the proportion of sequences, among all sequences with a DNA base (A, C, G, T) or gap (-) that has the majority consensus base.

The IUPAC consensus character includes all DNA bases that occur at a relative frequency higher than a user-specified ambiguity threshold. The threshold can range from 0 to 0.2: a value of 0 (default) will catch all the variation, but with the potential downside of generating oligos with high degeneracy. A higher value will capture most of the variation, but with the potential downside of missing less common sequence variants.

Coverage represents the proportion of sequences in the target alignment (among all sequences with a DNA base) that are covered by the IUPAC consensus character.

consensusProfile() has two arguments:

• x: a MultipleDNAAlignment object (see Data import)
• ambiguityThreshold: position wise “detection level” for ambiguous bases. All DNA bases that occur with a proportion higher than the specified value will be included in the IUPAC consensus character. Defaults to 0, can range from 0 to 0.2

Type the following code to generate a consensus profile from the example alignment below:

myConsensusProfile <- consensusProfile(myAlignment, ambiguityThreshold = 0.05)

Results (row 100-105):

The results can be visualized by using plotData():

plotData(myConsensusProfile)
#> Warning in ggplot2::geom_segment(color = "#93A8AC", ggplot2::aes(x = min(position), : All aesthetics have length 1, but the data has 7597 rows.
#> ℹ Please consider using `annotate()` or provide this layer with data containing
#>   a single row.

The dots represent the value at each position, and the black lines show centered running averages. High identity values (in combination with low gap values) indicate high sequence conservation.

The data can be explored in more detail by zooming into a specific region of interest:

## Select position 5000 to 5800 in the consensus profile 
selection <- myConsensusProfile[
  myConsensusProfile$position >= 5000 & myConsensusProfile$position <= 5800, 
  ]

plotData(selection)
#> Warning in ggplot2::geom_segment(color = "#93A8AC", ggplot2::aes(x = min(position), : All aesthetics have length 1, but the data has 801 rows.
#> ℹ Please consider using `annotate()` or provide this layer with data containing
#>   a single row.

Step 2: designOligos

The next step is to design oligos. Primers must be designed, and probes are optional. Primers can be generated by using two strategies:

• The ambiguous strategy (default) generates primers from the IUPAC consensus sequence alone, which means that ambiguous bases can occur at any position in the primer

• The mixed strategy generates primers from both the majority and IUPAC consensus sequence. These primers consist of a shorter degenerate part at the 3’ end (~1/3 of the primer, targeting a conserved region) and a longer consensus part at the 5’ end (~2/3 of the primer), which instead of having ambiguous bases, contains the most frequently occuring nucleotide at each position. This strategy resembles the widely used Consensus-Degenerate Hybrid Oligonucleotide Primer (CODEHOP) principle [2]. It aims to allow amplification of highly variable targets using primers with low degeneracy. The idea is that the degenerate 3’ end will bind specifically to the target sequence in the initial PCR cycles and promote amplification despite eventual mismatches at the 5’ consensus part (since 5’ end mismatches are generally less detrimental than 3’ end mismatches) [2]. In this way, the generated products will perfectly match the 5’ ends of all primers, allowing them to be efficiently amplified in later PCR cycles. To provide a sufficiently high melting temperature (tm) with eventual 5’ end mismatches, it is recommended to design relatively long primers (at least 25 bases) when using this strategy

Probes are always designed using the ambiguous strategy.

designOligos() has several arguments (design settings):

• x: the consensus profile from Step 1
• maxGapFrequency: maximum gap frequency (in the alignment) for primers and probes, defaults to 0.01
• lengthPrimer: primer length range, defaults to c(18, 22)
• maxDegeneracyPrimer: maximum degeneracy for primers, defaults to 4
• gcClampPrimer: if primers should have a GC clamp, defaults to TRUE. A GC clamp is identified as two to three G or C:s within the last five bases (3’ end) of the primer. The presence of a GC clamp is thought to promote specific binding of the 3’ end
• avoidThreeEndRunsPrimer: if primers with more than two runs of the same nucleotide at the terminal 3’ end should be avoided (to reduce the risk of mispriming), defaults to TRUE
• gcPrimer: GC-content range of primers, defaults to c(0.40, 0.65)
• tmPrimer: melting temperature range for primers (in Celsius degrees), defaults to c(55, 65). Melting temperatures are calculated for perfectly matching oligo-target duplexes using the nearest neighbor method [3], using the formula, salt correction method, and table values as described in [4]. See the manual (?oligos) for more information
• concPrimer: primer concentration in nM, defaults to 500 (for tm calculation)
• designStrategyPrimer: design strategy for primers, "ambiguous" or "mixed", defaults to "ambiguous"
• probe: if probes should be designed, defaults to TRUE
• lengthProbe: defaults to c(18, 22)
• maxDegeneracyProbe: defaults to 4
• avoidFiveEndGProbe: if probes with a G at the terminal 5’ end should be avoided (to prevent quenching of the 5’ flourophore of hydrolysis probes), defaults to TRUE
• gcProbe: defaults to c(0.40, 0.65)
• tmProbe: defaults to c(55, 70)
• concProbe: defaults to 250
• concNa: sodium ion (equivalent) concentration in the PCR reaction (in M), defaults to 0.05 (50 mM) (for calculation of tm and delta G)

All sequence variants of an oligo must fulfill all specified design constraints to be considered.

Oligos with at least one sequence variant containing more than four consecutive runs of the same nucleotide (e.g. “AAAAA”) and/or more than three consecutive runs of the same di-nucleotide (e.g. “TATATATA”) are not considered.

An error message will return if no oligos are found. If so, a good idea could be to re-run the process from Step 1 and increase the ambiguityThreshold in consensusProfile(), and/or relax the design constraints above.

myOligos <- designOligos(myConsensusProfile)

myOligos$sequence[[1]] ## All sequence variants of the first oligo (i.e., first row) 
#> [1] "TCCGCCCTGGCGAATGCTGT" "TCTGCCCTGGCGAATGCTGT" "TCCGCCTTGGCGAATGCTGT"
#> [4] "TCTGCCTTGGCGAATGCTGT"
myOligos$gcContent[[1]] ## GC-content of all variants of the first oligo 
#> [1] 0.65 0.60 0.60 0.55
myOligos$tm[[1]] ## Tm of all variants of the first oligo 
#> [1] 65.33624 62.86736 63.47792 61.00290

plotData(myOligos)

## I first need to make a consensus profile of the masked alignment 
myMaskedConsensusProfile <- consensusProfile(myMaskedAlignment, ambiguityThreshold = 0.05)

myMaskedMixedPrimers <- designOligos(myMaskedConsensusProfile,
                                     maxDegeneracyPrimer = 8,
                                     lengthPrimer = c(24:28),
                                     tmPrimer = c(65,70),
                                     designStrategyPrimer = "mixed", 
                                     probe = FALSE)

plotData(myMaskedMixedPrimers)

## Get the minimum score from myOligos 
bestOligoScore <- min(myOligos$score)
bestOligoScore
#> [1] 0

## Make a subset that only oligos with the best score are included 
oligoSelection <- myOligos[myOligos$score == bestOligoScore, ]

## Get the binding region of the first oligo in the selection above (first row): 
bindingRegion <- myConsensusProfile[
  myConsensusProfile$position >= oligoSelection[1, ]$start & 
    myConsensusProfile$position <= oligoSelection[1, ]$end,
  ]

plotData(bindingRegion, type = "nucleotide")

Step 3: designAssays

The final step is to find pairs of forward and reverse primers, and eventually, to combine them with probes. If probes are present in the input dataset, only assays with a probe located between the primer pair will be kept.

designAssays() has four arguments:

• x: the oligo dataset from Step 2
• length: amplicon length range, defaults to c(65, 120)
• tmDifferencePrimers: maximum allowed difference between the mean tm of the forward and reverse primer (in Celsius degrees). Defaults to NULL, which means that primers will be paired regardless of their tm. ¹

An error message will return if no assays are found.

Below, I show how to design assays using default settings using the dataset myOligos from Step 2:

myAssays <- designAssays(myOligos)  

Output (first five rows):

The output contains many columns, but you can get an overview by calling View(as.data.frame(myAssays)) if you are working in RStudio. Again, the results can be visualized using plotData():

plotData(myAssays)

There are now over 1000 assays, but we can reduce the number by filtering the dataset. In the example below, I select to subset the assays based on their average oligo score (see the score variable in the table above), and only keep those with the best score.

The best possible average score is 0 and the worst is 9 (see Step 2 for more information).

## Get the minimum (best) score from myAssays
bestAssayScore <- min(myAssays$score)
bestAssayScore
#> [1] 0.6666667

## Make a subset that only contains the assays with the best score 
myAssaySelection <- myAssays[myAssays$score == bestAssayScore, ]

Oligo and assay binding regions

Oligo and assay binding region(s) can be inspected in more detail by using the consensus profile from Step 1. As an example, I select to take a closer look at the first assay in myAssaySelection from Step 3.

The start and end position of an assay can be retrieved as follows:

from <- myAssaySelection[1, ]$start
to <- myAssaySelection[1, ]$end

It is now possible to indicate the amplicon region, using the highlight argument in plotData():

plotData(myConsensusProfile, highlight = c(from, to))
#> Warning in ggplot2::geom_segment(color = "#93A8AC", ggplot2::aes(x = min(position), : All aesthetics have length 1, but the data has 7597 rows.
#> ℹ Please consider using `annotate()` or provide this layer with data containing
#>   a single row.

To get a more detailed view, we can make a subset of the consensus profile to only contain the amplicon region, and plot it:

myAssayRegion <- myConsensusProfile[
  myConsensusProfile$position >= from & 
    myConsensusProfile$position <= to, 
] 

plotData(myAssayRegion, type = "nucleotide")

The consensus amplicon sequence is obtained as follows:

paste(myAssayRegion$iupac, collapse = "")
#> [1] "CRGTGGTTTCTGGGGTGACMGGGTTGATTCTCAGCCCTTCGCMMTCCCCTATWTTCATCCAACCAACCC"

Check match

It is often valuable to investigate how the generated oligos match with their target sequences. checkMatch() can be used for this purpose. The function is a wrapper to vcountPDict() from Biostrings [1] and has two arguments:

• x: an oligo or assay dataset from Step 2 or 3
• target: the target alignment used as input in Step 1

The output gives information on the proportion and names of target sequences that match perfectly and with one, two, three, or four or more mismatches to the oligo within the intended oligo binding region in the alignment (on-target match). It also tells the proportion and names of target sequences that match with a maximum of two mismatches to the oligo outside the intended oligo binding region (off-target match). Thus, be cautious that false negatives (or positives) may occur due to poorly aligned sequences. Moreover, the output does not tell which strand (minus or plus) the oligo matches to, which is important to consider when assessing off-target matches to single-stranded targets. See the manual (?checkMatch) for more information.

Ambiguous bases and gaps in the target sequences will be identified as mismatches.

checkMatch() can be used for both oligo and assay datasets. The function is rather slow, especially if there are many target sequences, so a good idea could be to select only a few oligo or assay candidates to investigate.

Below, I show how to check how the three first candidates of the oligoSelection dataset matches to the sequences in myAlignment:

## Check the first three candidates in the oligoSelection dataset
oligoSelectionMatch <- checkMatch(oligoSelection[1:3, ], target = myAlignment)

Output:

Note that the id variables can hold several values in each row. All values can be retrieved by:

## Get the id of all sequences that has one mismatch to the first oligo in the input dataset
oligoSelectionMatch$idOneMismatch[[1]] 
#> [1] "AB481228.1" "KJ701409.1" "JQ953665.1" "AB521805.1" "LC042232.1"

The output can be visualized using plotData():

plotData(oligoSelectionMatch)

Result tables

All result tables can be saved to .txt or .csv-files, for instance:

write.csv(myAssays, file = "myAssays.csv", quote = FALSE, row.names = FALSE) 
write.table(myAssays, file = "myAssays.txt", quote = FALSE, row.names = FALSE) 

Fasta-format

It is also possible to export oligos and assays in fasta-format. To do this, the object of interest must first be coerced to a DNAStringSet object, as follows:

## Convert the first two oligos 
as(myOligos[1:2, ], "DNAStringSet")
#> DNAStringSet object of length 6:
#>     width seq                                               names               
#> [1]    20 TCCGCCCTGGCGAATGCTGT                              oligo_1_variant_1
#> [2]    20 TCTGCCCTGGCGAATGCTGT                              oligo_1_variant_2
#> [3]    20 TCCGCCTTGGCGAATGCTGT                              oligo_1_variant_3
#> [4]    20 TCTGCCTTGGCGAATGCTGT                              oligo_1_variant_4
#> [5]    20 GCCCTGGCGAATGCTGTGGT                              oligo_2_variant_1
#> [6]    20 GCCTTGGCGAATGCTGTGGT                              oligo_2_variant_2

## Convert the first two assays 
as(myAssays[1:2, ], "DNAStringSet")
#> DNAStringSet object of length 12:
#>      width seq                                              names               
#>  [1]    20 GGCAGTGGTTTCTGGGGTGA                             assay_1_fwd_varia...
#>  [2]    20 GGCGGTGGTTTCTGGGGTGA                             assay_1_fwd_varia...
#>  [3]    20 GGCAGTGGTTTCTGGGGTGA                             assay_2_fwd_varia...
#>  [4]    20 GGCGGTGGTTTCTGGGGTGA                             assay_2_fwd_varia...
#>  [5]    20 CCTATATTCATCCAACCAAC                             assay_1_rev_varia...
#>  ...   ... ...
#>  [8]    20 CCTATTTTCATCCAACCAAC                             assay_2_rev_varia...
#>  [9]    18 CAGGGTTGATTCTCAGCC                               assay_1_pr_variant_1
#> [10]    18 CCGGGTTGATTCTCAGCC                               assay_1_pr_variant_2
#> [11]    19 CAGGGTTGATTCTCAGCCC                              assay_2_pr_variant_1
#> [12]    19 CCGGGTTGATTCTCAGCCC                              assay_2_pr_variant_2

Note that all sequences will be written in the same direction as the input target alignment, and all sequence variants of each oligo will be printed.

The sequences can now be saved in fasta-format by using writeXStringSet() from Biostrings [1].

toFile <- as(myOligos[1:2, ], "DNAStringSet")
writeXStringSet(toFile, file = "myOligos.txt")