Q: What is the cost of whole plasmid sequencing?

The cost is $15 per DNA sample and $20 per colony sample.

Q: What is the expected turnaround time?

For samples deposited in our dropboxes by 5 PM, you should expect sequencing results the following morning for DNA samples. Colony samples typically require an extra day for processing due to an additional miniprep step.

Q: Does Quintara process samples on the weekends?

All our sequencing services are available 7 days a week at our Boston facility and 6 days a week (excluding Sundays) at our California facility.

Q: What types of samples can I submit for nanopore sequencing?

We accept plasmid, RCA, amplicon (2kb+), and colony samples. For optimal results, we highly recommend submitting samples as either purified plasmids or colonies. These sample types have the highest reaction success rate. Colony samples are cultured and miniprepped prior to sequencing.

Q: What are the requirements for submitting plasmid samples?

For high-copy plasmids, we kindly request a minimum volume of 5 µL with DNA concentrations ranging between 100-200 ng/µL. We recommend submitting at least double the minimum volume so that we have enough DNA to repeat any reactions if necessary.

For low-copy plasmids, please provide us with the highest DNA concentration you can obtain from a miniprep with a minimum volume of 10 µL. This ensures we have the best chance of accurately sequencing your sample despite the lower amount of starting material.

Q: What are the requirements for submitting linear DNA samples (RCA/PCR)?

Please provide us with 10-20 µL of your undiluted sample. Giving us 20 µL ensures we have enough material to accurately sequence your DNA and perform repeats if necessary.

Prior to submission, please verify the quality of your PCR/RCA products by running them on a gel.

Though magnetic beads or column-based purification is not a mandatory step, we strongly encourage purifying your samples before submission as it significantly reduces the likelihood of reaction failure.

High concentrations are a must when sequencing linear DNA. We don’t recommend submitting your sample if you notice a dim band or multiple bands on the gel.

Q: What are the requirements for submitting colony/bacterial samples?

We accept colony samples in various formats: glycerol stock, agar plate, suspension in water/media, and bacterial culture.

To ensure that we accurately and efficiently process your agar plate orders, please carefully follow the instructions below:

1. Specify in Order Notes: Clearly indicate in your order notes that you are submitting multiple colonies from one or more agar plates. For example, if you have an agar plate labeled "Plate A" and wish to sequence 4 colonies from it, your note might say, "Sequence 4 colonies from Plate A."
2. Match Reactions to Colonies: The number of reactions you request should correspond to the number of colonies you want sequenced. This alignment is crucial for ensuring that we sequence each colony as intended.
3. Naming Conventions: When naming your reactions, incorporate the agar plate name along with a unique identifier for each colony. This method helps us precisely identify which colony corresponds to which sequencing reaction. For instance:
   • Plate A_colony1
   • Plate A_colony2
   • Plate A_colony3
   • Plate A_colony4

Q: Are colony samples sequenced directly?

We use your colony/bacterial sample to grow a culture followed by a miniprep prior to sequencing. Therefore, colony samples typically have a 2-3 day turnaround time.

While sequencing of colony samples using quicker methods is possible, these alternatives result in a higher failure rate and lower quality data. Purified plasmid DNA consistently yields the most accurate and reliable sequencing data. By starting with purified plasmid, we significantly reduce the risk of sequencing errors and reaction failures, ensuring that your data is both high-quality and dependable.

Q: Can I receive my leftover minipreps from the colony order?

Yes, we offer the option to return the remaining miniprep samples after sequencing. There is a delivery fee of $10 per order for this service. If you're interested in having your samples returned, please check the appropriate box on the order submission form.

Q: What if I don’t have enough DNA to meet the minimum sample requirement?

If your sample is a high-copy plasmid with a clean gel band and minimal genomic DNA contamination, concentrations as low as 20 ng/µL may still sequence well.

While nanopore sequencing can often succeed even at lower DNA concentrations, adhering to our recommended guidelines maximizes the chances of a successful reaction outcome. Please note that Quintara cannot be held responsible for reaction failures attributable to insufficient sample quality or quantity.

Q: What kind of tubes should I use for sample submission?

For sample submission, we highly recommend 8-well strip tubes or a 96-well PCR plate for those with 48 or more samples. The use of 1.5mL microfuge tubes is strongly discouraged due to the increased processing time they require and the higher risk of sample mix-ups, which could lead to delays in receiving your results.

Q: How do I place an order?

After logging into your Quintara account, please find the black panel located on the left side of the main dashboard. You should see “Whole Plasmid Sequencing” as an option under the “Nanopore” section. Click this to access the order form.

In the order form, please indicate the sample type, and fill in the mandatory sections (sample name, copy number for plasmids, amplicon size for PCR products, and culture type/antibiotics for colony samples). Other sections are optional (concentration, notes).

Q: Who do I contact if I have questions about my order?>

If you have any questions or need further assistance with your order, our technical support team is here to help. Please reach out to us at nanopore@quintarabio.com for prompt and detailed support.

Q: What is Quintara’s sequencing rerun policy?

If you encounter any issues with your sample results, do not hesitate to contact us at nanopore@quintarabio.com to discuss a rerun. For non-RCA samples, We offer one reaction repeat free of charge for any results that don't meet your expectations. Currently there is a basecaller problem with RCA samples that results in high reaction failure rates (currently being investigated by ONT). We therefore only resequence RCA samples on a case by case basis.

Q: When should I choose nanopore over sanger sequencing?

Nanopore sequencing is the preferred choice in several scenarios:

• Large, Complex Regions: If you're sequencing a region that requires three or more primers to cover effectively.
• Challenging Sequences: For regions of high GC content or strong secondary structures that are difficult for Sanger sequencing to navigate.

More details are available in the technical questions section of the FAQ.

Q: When should I choose sanger sequencing over nanopore?

While nanopore sequencing offers many advantages, there are instances where Sanger sequencing may still be more suitable:

• Small Regions: For sequencing small regions of interest, Sanger sequencing can be more cost-effective.
• Homopolymer Regions: Sanger sequencing is objectively better at resolving long homopolymer stretches (8bp or more), such as polyA tails, with higher accuracy.

Q: Does Quintara offer any discounts?

We do offer discounts based on order volume. Please email us at nanopore@quintarabio.com for further information!

Q: What is the maximum DNA length that can be effectively sequenced using this technology?

Our sequencing workflow is optimally designed to handle plasmids up to 30 kb in size.

For plasmids larger than 30 kb, it's important to note that while sequencing can still be performed, problems often occur during assembly of the consensus sequence. However, individual FASTQ reads can still be aligned to your reference map. This enables you to nanopore sequence larger plasmids, although with some considerations for data interpretation.

Q: Do you accept nanopore projects beyond whole plasmid sequencing?

Absolutely! Our team is actively working on developing new nanopore services to expand the range of applications and projects we support.

For custom projects that leverage the unique capabilities of the nanopore sequencing platform, we encourage you to contact us at nanopore@quintarabio.com to discuss your project ideas and how we can tailor our services to meet your needs.

Q: How can I provide feedback on your services?>

We highly value your input. For any feedback or suggestions, please email us at nanopore@quintarabio.com. Your insights help us serve you better.

Q: Can I have the raw reads associated with my order?

Certainly! This is not a typical request, so please email us directly at nanopore@quintarabio.com to receive your raw reads in FASTQ file format.

Q: Can I also have the FAST5/POD5 files associated with my order?

We can certainly provide those as well, although this is an even less common request. Please reach out to us at nanopore@quintarabio.com.

Q: How does whole plasmid (nanopore) sequencing work?

Whole plasmid sequencing is performed by tiny protein channels known as nanopores. As a DNA strand passes through a nanopore, it disrupts an electrical current. These current disruptions are picked up by a sensor and translated into a DNA sequence. Each DNA strand generates a read representing the sequence of a single molecule in your sample. By sequencing numerous molecules and pooling the data, we can reconstruct the complete sequence of your plasmid.

Q: What kinds of data are included in my results?

Your sequencing results are organized into folders, each dedicated to a specific type of data.

Q: What sequence files can I use to align directly against my reference sequence?

For aligning sequence data, you can utilize the following files:

• Chromatogram (ab1) Files: These offer a Sanger-like representation of your nanopore data.
• FASTA Files: A standard text-based format containing the raw sequence data, ready for direct alignment tasks.
• Genbank Files: Annotated with plasmid features, these files not only provide sequence data but also contextual information about the plasmid, which can be particularly useful for visualization and alignment in tools like Snapgene Viewer.

Q: Which sequence file SHOULD I use for alignment purposes?

All three file types—ab1, FASTA, and Genbank—are suitable for alignment tasks. However, we prefer the ab1 file since it also includes detailed per-base information. For example, positions of low confidence are identified as heterozygous (mixed) peaks, akin to Sanger mixed peaks and low coverage positions appear as shorter peaks with heights proportional to coverage depth.

It's important to note that the ab1 format is artificially generated since nanopore sequencing does not produce actual chromatogram data. Consequently, not all alignment software support this file type. If you encounter compatibility issues, we suggest using Snapgene Viewer, which is available for free download.

Q: Why are your sequence files labeled with the word “contig”? What does this mean?

The term "contig" refers to a consensus sequence file generated by our whole plasmid sequencing pipeline. This label signifies that the sequence you're viewing is constructed from numerous reads aligned and merged to form a single, consensus sequence.

When your sample contains multiple DNA species, our pipeline may produce several contigs. These are systematically numbered to differentiate between them.

By labeling these sequences as "contigs," we aim to provide a clear indication of their origin.

Q: What are the advantages of nanopore over sanger sequencing?

Nanopore sequencing offers a range of advantages that address some of the limitations encountered with traditional Sanger sequencing. Key advantages include:

• Lower Failure Rate: Nanopore sequencing exhibits a significantly reduced failure rate for plasmid samples.
• No Primers Required: This technology is primer-free, simplifying the preparation process and reducing the overall cost and time involved in sequencing.
• Cost-Effective for Unknown Plasmids: Since primers aren’t required, nanopore sequencing is an effective way to verify unknown plasmids.
• Complex Sequence Compatibility: It excels at sequencing through challenging regions such as GC-rich sequences, repetitive areas (excluding homopolymers), and strong secondary structures like hairpins and inverted terminal repeats (ITRs), where Sanger sequencing often struggles.
• Sample Quality Insights: Nanopore technology can also provide valuable sample quality information, including the percentage of E-coli genomic DNA contamination, sample purity (identification of other DNA species), and multimerization.

Q: What are the disadvantages of nanopore over sanger sequencing?

While nanopore sequencing offers numerous advantages, it also has a few, specific limitations when compared to traditional Sanger sequencing:

• Long Homopolymers: Nanopore sequencing faces challenges in accurately sequencing long homopolymer regions (sequences of the same nucleotide repeated more than 8 times). These regions can sometimes result in insertion-deletion (indel) artifacts.
• Artifacts in Methylated GATC Regions: Occasional sequencing artifacts can occur in GATC sequences, caused by dam methylation.

We expect these issues to become less relevant as the technology progresses.

Q: What is multimerization?

Multimerization refers to the process whereby copies of the same plasmid within a bacterial cell undergo recombination, leading to the formation of concatenated plasmids. These can range from dimers and trimers to tetramers and beyond, resulting in a portion or all of the sample consisting of these multi-copy plasmids. This phenomenon came to light once long-read sequencing became more widely used.

Multimers are undetectable using Sanger sequencing. They're also hard to identify with gel electrophoresis since digesting a multimer would cause it to resemble its monomer form. For a deeper understanding of plasmid multimerization, check out the Addgene article: Plasmids 101: Dimers and Multimers (addgene.org)

Q: How accurate are the sequencing results?

Outside of dam methylated and homopolymer regions, nanopore sequencing is very accurate; we almost never see any artifacts provided there is sufficient depth of coverage.

If you encounter any unexpected or questionable results, or suspect a mis-assembly of your plasmid sequence, we encourage you to contact us directly at nanopore@quintarabio.com. We will perform a deeper analysis of the results to determine whether there are any potential artifacts present in the data.

Q: How much coverage do I get per sample?

We aim for 500+ reads per sample, but this can vary greatly depending on sample quality, plasmid copy-number, sample type, and even what samples from other customers are being run alongside yours. If the sample quality is good, we can typically achieve accurate results with fewer than 200 reads.

Q: IMPORTANT: HOW DO I ANALYZE THE PDF REPORT?

We strongly urge all our clients to examine the report file that comes with each sequencing reaction. The information contained within could save you countless hours of time.

The report is broken down into several sections:

Section 1 - Sample Name and Processing Date:

This section displays the sample name along with the date on which the data was processed.

Section 2 - Read Count Table:

This table contains information about the total number of reads obtained from the sequencing reaction as well as the percentage of host genomic DNA in the sample.

In the example table below, a total of 1,673 reads were obtained for this sample, corresponding to 6,314,277 bp of sequence data. This latter number is derived from the sum of all reads multiplied by their respective read lengths.

For this example, reads originating from the host genome comprise 0.48% of all reads and 0.73% of all bp sequenced.

When estimating the amount of genomic contamination within your sample, we recommend using the percentage of total bases rather than the percentage of total reads.

Note that shorter (highly fragmented) E-coli genomic reads may not be properly categorized. This is because setting more aggressive filtering conditions could cause actual plasmid reads to be wrongly categorized as genomic DNA. We advise you treat the genomic contamination values as an underestimate.

Section 3 - Contig Details Table:

This table contains information about each consensus (contig) sequence generated by the assembly pipeline. Details include sample name, sequence length, the percentage and number of reads that map to that particular contig, and whether the contig is circular.

In the example below, there is one consensus sequence with a length of 5431 bp. 1442 reads mapped to it. The percentage of bases mapped is 95% and the consensus represents a circular DNA (plasmid).

You’ll notice that 95% of bases mapped to the consensus. In other words, 5% didn't map. The missing 5% are likely E-coli genomic reads that weren't properly filtered out or could be some unknown DNA.

Although the assembly software is able to classify and filter out most genomic DNA reads, there is a minimum read-length cutoff to ensure that reads from non-genomic sources aren’t accidentally misclassified as genomic DNA. As a result, a certain percentage of shorter genomic DNA reads aren’t properly categorized. Of course there may also be some reads that are truly of an unknown origin.

If your sample consists of one species, you should expect to see a single contig displayed on this table. If multiple contigs are listed, there is a good chance that your sample contains multiple DNA species.

For a pure plasmid sample, you should expect the mapping percentage to be greater than 80%. Percentages below that tend to correlate with high levels of genomic DNA in the sample.

If coverage depth is inadequate, sometimes the assembly pipeline will miscategorize a circular DNA as linear.

It should be noted that even if multiple species are present in your sample, the assembly software may not detect all of them, or the assembly could fail as a result. It’s important to scrutinize the read-length histogram to figure out if multiple species are indeed present (details on this in a later section).

Section - 4: Coverage Map:

The coverage map is a graph that plots the coverage depth (the number of reads that map to each base of your consensus sequence) on the Y-axis and the base position of the consensus sequence on the X-axis. You should look for any sudden, sharp changes in read depth as that’s a possible indication that there are several related species within your sample.

The base position axis on the graph is relative to your sequence file, NOT your reference map! For example, if you see a sharp drop in read depth from positions 2000-3000, you should open the genbank or FASTA sequence file and analyze the sequence from base positions 2000-3000. DO NOT open your own reference map and look at positions 2000-3000.

The coverage map is a great tool for figuring out if the contaminant species within your sample are derived from or related to the consensus sequence. For example, if you have a 10 kb plasmid and somehow 20% of the plasmid population lost 2kb of the insert during the culture process, you should see a sharp 20% drop in coverage depth spanning 2 kb around the region of the sequence where the deletion occurred.

Any low confidence positions will be marked with an orange “x” on the coverage map.

Section 5 - Read-Length Distribution:

The read-length distribution displays the number of reads of each read-length used for the contig assembly.

This is by far the most important section of the entire pdf report. You should at the very least check the read-length distribution for all of your samples.

Disclaimer: The read-length distribution is only useful for circular DNA samples. You should not use it to draw conclusions about sample purity of linear DNA (RCA/PCR). This is because any linear DNA will be fragmented during the library prep process, resulting in a wide range of peaks in the read-length distribution. This makes it indistinguishable from indicators of potential contamination.

For clean plasmid samples, you should see either a single tall peak or two adjacent tall peaks in the graph. Examples below indicate pure samples:

The read-length(s) of the tall peak(s) should be roughly equal to the expected size of your plasmid.

If you come across multiple non-adjacent tall peaks in the read-length distribution, one of two scenarios likely applies to your sample. The first is somewhat benign while the second requires your immediate attention.

Q: Can I use the read-length distribution to analyze non-plasmid samples?

Nope. This is because any linear DNA will be fragmented during the library prep process, and this is captured as a wide range of peaks in the distribution, which is indistinguishable from indicators of potential contamination. The read-length histogram is only valid for determining sample purity for CIRCULAR DNA (i.e. plasmids, but circular viral DNA may work as well).

Q: I noticed there’s sometimes two adjacent tall peaks in my read-length distribution. Is this an indication of contamination?

Your sample is fine. In reality, most of the reads from your sample will lose a couple bases from the ends during the library prep process. As a result, read lengths typically fall within a size range. The read-length distribution generated in the report may have different bin widths from sample to sample and this causes the plasmid peak to appear as either a single peak (if the histogram bin width is large), or two adjacent peaks (if the bin width is small). However, please make sure the two peaks are NOT separated by one or more bin widths. Otherwise that is very likely an indication that your sample contains two different plasmids of similar sizes.

Q: What is considered a high level of genomic DNA contamination?

We’ve noticed that the extent of genomic DNA contamination can depend on the process used to purify the plasmids. Cesium chloride-based purification methods produce DNA with the lowest levels of E-coli genomic contamination (~0%). Silica column preps (typically minipreps) also produce relatively clean preps (0.1-2% genomic DNA contamination). Anion exchange preps typically have higher levels of E-coli contamination (1-20%), though this could be due to longer culture times during scale-up.

Copy-number also impacts the level of genomic contamination. Low-copy plasmids tend to have very high levels of genomic contamination while high-copy plasmids are much lower.

Most plasmid samples fall within the 0-20% genomic contamination range. If your sample is higher, that may be a cause for concern.

Q: I received more than one sequence (contig) for my sample. What does this mean?

If your sample is supposed to be a plasmid, then there's probably some other contaminant present, likely another plasmid species. We recommend analyzing the read-length distribution in the report pdf to check for peaks corresponding to each consensus sequence. If each consensus has a corresponding peak in the histogram, that's a strong indication that a contaminant DNA species is present.

If your sample is linear DNA (RCA/PCR), it’s more likely that the additional sequences you received are assembly artifacts generated during downstream processing. For linear DNA, assembly is much more challenging and oftentimes will generate multiple sequence contigs even if the sample contains only one species. We advise that you ignore all the contigs that don’t match your reference sequence. Typically, the longest contig is the correct one. At present it's not possible to draw any conclusions about sample purity for linear DNA samples.

Q: I see multiple peaks in my read-length distribution. Does that mean there may be contaminants?

If there are peaks with sizes corresponding to double, triple, or even quadruple the expected size of your plasmid, your sample is likely affected by multimerization.

If your sample is not a purified plasmid, the read-length distribution can’t be used for purity analysis.

If you see two adjacent peaks with sizes roughly corresponding to the expected size of your plasmid, that’s a sign your sample has no issues.

Otherwise, yes. Your sample probably is contaminated.

Q: How reliable and accurate is the information displayed in the read-length distribution?

These results are highly reliable, virtually on par with gel electrophoresis for determining DNA length. Here’s how it’s measured: the DNA strand is guided by a motor protein when passing through the nanopore and always travels at a constant rate of 400 bp/second. Taking this information and the time it takes for the DNA strand to fully pass through the nanopore (which is measured by the sensor), we can accurately determine the length of the DNA strand/read.

There are two main caveats:

The sample MUST be a circular plasmid.

Even if the sample IS a circular plasmid, there can’t be too many peaks in the read-length distribution, which is an indication that there’s a lot of highly-fragmented genomic DNA present that weren't properly filtered out. This usually correlates with a high percentage of genomic DNA in the Read Count Table of the PDF report.

Q: Can I sequence my sample if I know it contains a mixture of plasmids, amplicons, etc?

This is not recommended as the assembly workflow is geared towards clonal samples. Sometimes, the assembly software will correctly identify some if not all of the species within a mixture, but the reaction is more likely to fail due to the inability to reconcile all the disparate sequences originating from different sources. If you're troubleshooting to figure out if something is wrong with your sample, the read-length distribution would still be very useful even if your reaction failed.

Q: My plasmid contains very large repeats (1kb+). Would this affect contig assembly?

Assembly works fine for repeats less than 1 kb. However for larger repeats, especially very large ones (4kb+), problems may occur during the assembly process. If your sample contains a large repeat, please reach out to us at nanopore@quintarabio.com so that we can apply special conditions to your reactions during the assembly step.

Q: The length of my consensus sequence doesn’t match up with the tall peak’s corresponding read length in the distribution. What does this discrepancy mean?

This section applies only to plasmid samples.

Assuming your sample is a plasmid, you should first check if the peak in the read-length distribution is a multiple of the size of your consensus sequence. If so, then most likely the majority/all of the DNA in your sample have multimerized.

Otherwise, there are two points to consider:

The consensus sequence is derived from a complex computational process that assembles together many reads. The process is not entirely foolproof. Assembly errors do happen from time to time, though it is typically quite rare for clonal plasmid samples.

On the other hand, the read-length distribution is derived from a precise measurement made during the sequencing process and is highly reliable. The read-length distribution is not determined by the assembly process.

Given these two points, if there is a discrepancy between the consensus sequence and read-length distribution, you should always trust the read-length distribution. You can also reach out to us and ask for a re-assembly of your data. Sometimes, just reprocessing your data will produce the correct consensus sequence.

Q: Why are there ambiguous (non-ATCG) bases in my ab1 (chromatogram) file?

When designing the chromatogram file, we made the conscious choice to mark any heterozygous/non-clonal peaks with an ambiguous base so that it’s easier to catch during sequence alignment.

A heterozygous peak occurs when a significant proportion of reads map to two or more unique nucleotides at a specific position in the consensus sequence. For example, if 70% of reads map to a C and 30% of reads map to a T at position 2182 of the consensus sequence, you will see a “mixed” peak consisting of a taller C (blue) peak and a shorter T (red) peak. You will also notice the base is called as a “Y” instead of a “C”. The “Y” allows for easier detection during analysis and represents the ambiguity of a C + T.

Q: Why are there sometimes multiple ab1 files associated with a single contig? Should there be any overlap between the ab1 files?

Chromatogram (ab1) files are designed for Sanger sequencing, which generates reads up to only 1000 bp in length. Due to this inherent limitation in the ab1 file format, we must divide longer sequences, such as full plasmid sequences, into smaller segments. This division is necessary to accommodate the file type's read-length restriction. Consequently, each segment should be distinct, without overlap in the traces. If you observe overlapping between nanopore ab1 traces, that indicates the presence of an insertion in your contig sequence relative to your reference.

Q: I notice mismatches/low confidence positions outside of homopolymer regions. Are these real or artifactual?

If the coverage depth is low (<100), the mismatches could be real or artifactual; it's difficult to tell.

For linear DNA samples, artifacts may occur at the ends of the consensus sequence even with high coverage depth.

For plasmids, there is a frequent occurrence of a particular type of artifact:

If mismatches are occurring on or immediately adjacent to the “GATC” sequence pattern, these are almost always artifacts caused by dam methylation. These artifacts follow a very distinct pattern which make them easy to spot:

The C base at the end may be miscalled as a T or there may be a mixture of C + T (indicated as an ambiguous Y base in the ab1 file).

Q: How do I analyze the per-base CSV data?

Most of the information found in these CSV files are already incorporated into the ab1 file type, so if you’re using ab1 files for sequence alignment, there isn’t really much of a reason to check the per-base data.

The “per_base_details” CSV file contains a table with each row corresponding to a specific base of the contig sequence.

For example:

Below is a description of the information found in each column (numbered from 1-13):

1. pos: indicates the base position within the contig sequence.
2. base: identifies the nucleotide base that is most likely to be correct.
3. depth: the number of reads that map to the base position
4. match: the number of reads that match the likely correct base from column 2.
5. vaf: refers to the “Variant Allele Frequency” - the proportion of total reads that map to the likely correct base. This is calculated by dividing the match value (column 4) by the depth value (column 3).
6. G: the number of reads that map to the G base.
7. A: the number of reads that map to the A base.
8. T: the number of reads that map to the T base.
9. C: the number of reads that map to the C base.
10. ins: the number of reads that map to a point insertion.
11. del: the number of reads that map to a point deletion.
12. homo: indicates whether the base position is part of a homopolymer stretch and identifies the homopolymer base.
13. confidence: marks any low-confidence base position with the word “low”.
14. In a separate “low_confidence_bases” CSV are all the corresponding base positions marked as low confidence in column 13 of the “per_base_details” CSV.