Coffalyser.Net™ uses colours and symbols to draw attention to deviating probe ratios. This article provides an overview of the most common colours and symbols that may be encountered.

The colouring system may seem complex at first. It is usually sufficient to remember that any non-standard colouring applied to a probe ratio is an indication that the result requires additional inspection.

The most important criteria used by Coffalyser.Net to decide on a colour are as follows:

• Whether the probe has a signal.
• Whether the signal is significant (> 10% of the median signal of the reference probes).
• Whether the signal is significantly different from the reference samples.
• Whether the signal crosses one of the arbitrary borders.

The table below lists the most common situations for most probe types. Different formats are used in PDF reports, grids and ratio charts due to the different nature of these views. The table below lists the colouring and symbols used in PDF reports, the colours used in grids within the software, and the symbols used in ratio charts. Click on an symbol to show or hide additional information about a situation.

Situation	Arbitrary Borders	PDF Font	PDF [REF]	Grid Colour	Ratio Chart
No significant difference	not crossed	font	=	× × × × ×
A probe ratio that is not significantly different from the reference samples and does not cross one of the arbitrary borders does not indicate a copy number change. The probe ratio is displayed in a neutral style. In grids, the colour of the cells may vary depending on the region the probe is in, and on whether the probe is a reference probe or a target probe. In ratio charts, the 95% confidence interval estimate of the probe (error bars) will overlap with the 95% confidence interval of the probe over the reference samples (box in the background).
Decreased by 2 SDs or more	not crossed	font	<<	×
A probe ratio that is significantly lower than in the reference samples (more than 2 standard deviations) but that does not cross the lower arbitrary border is indicated in this style. For a probe that targets a sequence with a normal copy number of 2, this is usually an unexpected result that may indicate things like experimental issues or mosaicism. For a probe that targets more than two copies per cell, a result like this may be expected, e.g. in case one of these copies is deleted. For example, a probe that targets a sequence with a normal copy number of 4 may show a ratio of ~0.75 when one of the four copies is deleted. In ratio charts, the 95% confidence interval estimate (error bars) does not overlap with the 95% confidence interval of the probe over the reference samples (box in the background). In PDF reports, the symbol << is used to indicate that the probe is at least two standard deviations lower than in the reference samples.
Increased by 2 SDs or more	not crossed	font	>>	×
A probe ratio that is significantly higher than in the reference samples (more than 2 standard deviations) but that does not cross the upper arbitrary border is indicated in this style. For a probe that targets a sequence with a normal copy number of 2, this is usually an unexpected result that may indicate things like experimental issues or mosaicism. For a probe that targets more than two copies per cell, a result like this may be expected, e.g. in case one of these copies is duplicated. For example, a probe that targets a sequence with a normal copy number of 4 may show a ratio of ~1.25 when one of the four copies is duplicated. In ratio charts, the 95% confidence interval estimate (error bars) does not overlap with the 95% confidence interval of the probe over the reference samples (box in the background). In PDF reports, the symbol << is used to indicate that the probe is at least two standard deviations lower than in the reference samples.
Decreased by 2 SDs or more	crossed	font	<<*	×
A probe ratio that is significantly different from the reference samples (more than 2 standard deviations) and that crosses the lower arbitrary border is indicated in red. This may, for example, be a heterozygous deletion of a sequence that has a normal copy number of 2 in the reference samples (final ratio ~0.5). In ratio charts, the 95% confidence interval estimate (error bars) does not overlap with the 95% confidence interval of the probe over the reference samples (box in the background). In PDF reports, the symbol <<* is used to indicate that the probe is at least two standard deviations lower than in the reference samples (<<) and that it crossed the arbitrary border (*).
Increased by 2 SDs or more	crossed	font	>>*	×
A probe ratio that is significantly different from the reference samples (more than 2 standard deviations) and that crosses the upper arbitrary border is indicated in blue. This may, for example, be a heterozygous duplication of a sequence that has a normal copy number of 2 in the reference samples (final ratio ~1.5). In ratio charts, the 95% confidence interval estimate (error bars) does not overlap with the 95% confidence interval of the probe over the reference samples (box in the background). In PDF reports, the symbol >>* is used to indicate that the probe is at least two standard deviations higher than in the reference samples (>>) and that it crossed the arbitrary border (*).
Decreased by 1-2 SDs (not significant)	crossed	font	<*	×
This style is used for a probe ratio that is lower than in the reference samples, but not significantly so (between 1 and 2 standard deviations), while it does cross the lower arbitrary border. This may, for example, be due to high variability in the experiment. In ratio charts, the 95% confidence interval estimate (error bars) overlaps with those of the reference samples (box in the background). In PDF reports, the symbol <* is used to indicate that the probe is only one standard deviation lower than in the reference samples (<) and that it crossed the arbitrary border (*).
Increased by 1-2 SDs (not significant)	crossed	font	>*	×
This style is used for a probe ratio that is higher than in the reference samples, but not significantly so (between 1 and 2 standard deviations), while it does cross the upper arbitrary border. This may, for example, be due to high variability in the experiment. In ratio charts, the 95% confidence interval estimate (error bars) overlaps with those of the reference samples (box in the background). In PDF reports, the symbol >* is used to indicate that the probe is only one standard deviation higher than in the reference samples (>) and that it crossed the arbitrary border (*).
Ambiguous	crossed	font	?	×
A probe that is less than 1 standard deviation different from the reference samples but that did cross the lower or upper arbitrary border is displayed in this style. This may, for example, be due to very high variability or due to other experimental issues such as an unsuitable reference sample selection. In PDF reports, the symbol ? is used to indicate that the result is unclear.
Mutation-specific probe with signal > 10% of the median signal of the reference probes	n/a	font	m	×
A mutation-specific probe with a significant signal (> 10% of the median signal of the reference probes in the reaction) is shown in a blue style to indicate that its signal is "higher than expect" (because no signal is expected for a mutation-specific probe in a normal sample). The result of a mutation-specific probe is not shown as final ratio but as intra ratio percentage (more details), which means that orange boxes instead of dots are used in ratio charts. In PDF reports, the symbol m is used to indicate that a mutation-specific signal has been detected. Mutation-specific probes with a signal ≤ 10% of the median signal of the reference probes or with no detected signal are shown in the same style as other probes (see rows below).
Signal without signal in most reference samples	n/a	font	INF	×
If a probe has a signal but the majority of the reference samples does not, the probe ratio should technically be infinite. For a probe like this, a ratio is calculated based on the reference samples that do have a signal. As this is dangerous, such a probe is clearly indicated in orange. In PDF reports, the probe is indicated with INF (for infinity). This situation is not common, as reference samples should have signals for all probes (except mutation-specific probes) in most cases. This situation is, therefore, most often encountered in special probemixes, when no dedicated reference samples are used, and/or when incorrect reference samples have been selected. Read more about the importance of reference samples.
Signal ≤ 10% of the median signal of the reference probes	n/a	font	variable	×
A very low signal ≤ 10% of the median signal of the reference probes is unexpected in most cases. Because it may actually be a background signal or noise peak, a special style is used. The result is also shown as intra ratio percentage, as the calculation of a final ratio using very low values is difficult (more information). In PDF reports, various symbols may be shown depending on the expected result for the probe and whether the reference samples had a (similar) signal.
No signal	n/a	font	<<**	×
A probe without a signal is indicated as "missing" using a red font and bright yellow cells in the grids. In PDF reports, the symbol <<** is used to indicate that the signal is absent.