By default Spectronaut™ will suggest a list of differently abundant candidates from your analysis. That's because Differential Abundance Testing is set to Student's t‑test on the Post Analysis node ‑under the BGS Factory Settings (default)‑ (figure 1). Spectronaut will assess candidates based on two widely used criteria: 1) the level of statistical significance on a Student's t‑test strategy and 2) the value of the fold change between the two conditions compared.
Figure 1. Candidates will be assessed only if Student's t‑test is selected on the Post Analysis node of the Analysis settings
1. To check for significant changes, Spectronaut runs one sample Student's t‑tests on the log ratios of the quantitative data points between every two conditions. These data points are the quantified precursors. Noteworthy, t‑tests are done for all precursors, even if a Top N strategy (or proteotypicity) is selected for quantification. The obtained p-value space is then corrected for multiple‑testing, following the method described by Storey (2002); this correction results in the provided q-values. To know more about this q-value, please keep on reading here.
2. The fold changes are calculated by averaging the replicates of each condition and performing a ratio of means. In contrast to the q-values, the fold changes are calculated using only those precursors or peptides used for quantification (fulfilling the Top N strategy and the proteotypicity, if selected).
These two features are shown on the Post Analysis Perspective, under the Differential Abundance node in the Candidates table (figure 2). The fold changes are log2 transformed and presented under the header AVG Log2 Ratio. The candidate list is by default filtered for a minimum q-value of 0.05 and a minimum (absolute) log2 ratio of 0.58. You can change these thresholds by clicking on the filter field and writing your preferred value (figure 2). These changes on the candidate list will apply as well on the Volcano Plot.
Figure 2. Candidates table. You can change the default threshold by clicking on the header and writing your preferred cut‑off
On the Candidates table you can also find a column labeled as [DEPRECATED] AVG Log2 Ratios (and [DEPRECATED] Absolute AVG Log2 Ratios). These columns correspond to the fold changes the way they were calculated in the past (averaging all the ratios going to pair‑wise comparisons). They are not wrong, but we found the current method to be more accurate in our internal tests.
Storey JD (2002) A direct approach to false discovery rates. J R Stat Soc Ser B Stat Methodol 64:479–498.
Created by SEZ. Last update 2018-03-14 by SEZ