How many genes that have plant-like composition are present in fungal libraries? The answer varies with the degree of confidence that one chooses to assign to the inference. At 95% confidence (P < 0.05), 36 sequences in Harrison's library, 45 sequences in Lammers' library, and 16 sequences in Sawaki's library resemble plant training sequences in their hexamer composition. Reasons why this might be so are considered below.
How many transcripts in the plant root-hair enriched library appear fungal? The majority. Recall that the null hypothesis here is that a transcript is of fungal origin. In the case of plant sequences, we are unable to reject the null hypothesis for 89% of 899 sequences (P > 0.05). This yields 800 plant root-hair transcripts that do not significantly resemble plant training sequences in hexamer content! Interpretations of this result are discussed below.
How many transcripts are bacterial? No sequences in any of the libraries significantly resemble rhizobacteria in hexamer content (P > 0.05). Conversely, it may be worth asking whether some bacterial genes resemble eukaryotic training sets in future investigations.