Simon Andrews

With the introduction of the NextSeq system Illumina changed the way their image data was acquired so that instead of capturing 4 images per cycle they needed only 2. This speeds up image acquisition significantly but also introduces a problem where high quality calls for G bases can be made where there is actually no signal on the flowcell.

With the increasing capacity of a single flowcell lane it can be tempting to mix samples of different types within the same lane to make the most of your sequencing, but cross contamination between libraries in a flowcell can lead to the generation of artefacts which can mess up your analysis.

Probably the single biggest problem with the mapping of reads to a reference sequence is dealing with reads which come from parts of the genome which aren’t in the assembly. These reads can cause significant amounts of noise in anlayses performed on genomic data.

One of the standard fields in the SAM/BAM file format is the mapping quality (MAPQ) value. This value can be very useful to help filter mapped reads before doing downstream analysis – unfortunately the implementation of this value is in no way consistent between different aligners so it takes a fair bit of research to know how to use it appropriately. Mis-applying the filter could cause reads to be inappropriately excluded from an analysis.

In some experimental designs a large proportion of the sequences in a library can have identical sequence at their 5′ end. These types of library can cause problems for the data collection and base calling on illumina sequencers, leading to the generation of poor quality data.

We are increasingly re-using data deposited in public sequence archives such as SRA, ENA or DDBJ and we rely on being able to successfully extract data from these sources. In some cases we have found that errors in the validation of the data can mean that data is corrupted when it is downloaded from these repositories.

When running multiple samples in the same sequencing reaction the different libraries are usually created with unique sequence tags (barcodes) to allow the sequence from the lane to be separated. Problems during this splitting are common and can have serious effects on downstream analysis.

In a dataset where you have some degree of technical duplication, and have not filtered your data to only keep uniquely mapping reads then if you perform deduplication it will look as if repeat sequences are enriched.

Many sequencing platforms require the addition of specific adapter sequences to the end of the fragments to be sequenced. For an individual fragment, if the length of the sequencing read is longer than the fragment to be sequenced then the read will continue into the adapter sequence on the end. Unless it is removed this adapter sequence will cause problems for downstream mapping, assembly or other analysis.

The assumption when analysing sequence datasets is that every sequence comes from a different biological fragment in the original sample. Many library preparation techniques though include one or more PCR steps which introduce the possibility that the same original fragment can be observed multiple times, biasing the results produced. In some cases this type of duplication can be extreme and have a serious effect on the ability to analyse the data correctly.

For speed and efficiency many RNA-Seq mapping protocols map either entirely or initially to a transcriptome sequence. When the sample is contaminated with genomic material you can get significant numbers of reads reported as uniquely mapping when they actually map to many locations within the genome.

In theory RNA-Seq samples only contain reads derived from RNA. Most protocols include a DNase treatment to remove any remaining DNA from the sample. It is fairly common however to see samples with significant amounts of DNA contamination. Ignoring this can bias the counts generated from the data and throw off normalisation.

One of the biggest problems with sequencing libraries is that the material might be contaminated with something unexpected. One of the simplest forms of contamination is where you have material from a different species than expected. In many cases the rogue material will come from a species you can guess based on their other species commonly used in your lab. Screening for this type of contamination will help spot when you have contaminated samples, and can also help when you have completely switched samples.

The construction of sequencing libraries on many platforms requires the addition of specific adapter sequences to the ends of the fragments to be sequenced. Although there are steps in place to ensure that only valid adapter+insert combinations make it onto the sequencer it is possible to get adapter dimers with no valid insert making it through the sequencing process.

In a randomly primed library there is no reason to expect that a specific sequencing cycle should contain more of one specific base than any other cycle. It is commonly observed though that this type of library does contain a cycle-specific sequence bias, most frequently in the initial bases of the run.

Rather than a general loss of quality for a whole sequencing lane or run sometimes partial failures occur. These can affect specific regions and cycles and can have knock on effects for the data generated

Base call qualities in fastq and BAM format files are recorded using a text based encoding. Mistakes in the way this encoding is applied can lead to incorrect interpretation of base call accuracies in downstream analyses.

Sometimes a sequencing run can experience a sudden and lasting loss of base call quality across all sequences.

Illumina based sequencing shows a loss of base call quality as the number of sequencing cycles performed increases.