Outputting the results

The program NManout2 will allow us to output the results in various formats. First, let's use the flag option --summary to get a single file summarizing the ANOVA results as a table with the following columns:

1. the chromosome name,
2. the chromosome chunk index,
3. the starting position of the reference nucleosome (here BY),
4. the ending position of the reference nucleosome (here BY),
5. the starting position of the query nucleosome (here RM),
6. the ending position of the query nucleosome (here RM),
7. the p-value from the ANOVA for the interaction term (-1 for unscanned nucleosome),
8. the sign of the interaction (0 = reference, 1 = query),
9. the value of the interaction term,
10. the status of the reference nucleosome (1 = fuzzy, 2 = well-localized),
11. the status of the query nucleosome (1 = fuzzy, 2 = well-localized).

So let's generate this file:

$NManout2 -i S5/NManova -d HucH3K14_norm.db --summary 1>S5/NManova_summary.txt

Once it is done, you should end up with the following file:

$head S5/NManova_summary.txt
chrIII 1 2722 2850 2682 2818 2.78859e-14 0 1.26806 2 2
chrIII 1 2858 2978 2826 2986 9.98774e-18 0 1.4422 2 1
chrIII 1 3006 3138 3014 3146 2.53145e-26 0 1.14352 2 2
chrIII 1 3154 3274 3170 3326 5.93676e-07 0 0.879466 2 1
chrIII 1 3294 3430 3334 3454 8.5769e-18 0 1.42377 2 2
chrIII 1 3438 3558 3462 3626 1.4782e-24 0 1.57474 2 1
chrIII 1 3582 3738 3634 3798 6.06824e-25 0 1.48753 1 1
chrIII 1 3770 3922 3834 3962 2.89049e-07 0 0.840172 1 2
chrIII 1 4002 4138 3982 4134 4.31898e-06 0 1.00544 2 1
chrIII 1 4314 4466 4374 4498 0.032396 0 0.517827 1 2

where the meaning of the columns has been discussed above.

To get a more detailed picture of the SNEPs, we can then ask to NManout2 to output only the nucleosomes which p-values are lower a pre-defined cut-off together with additional information such that neighboring nucleosomes and known functional annotations. So, here we will look at the nucleosomes which p-values are smaller than the cut-off corresponding to a FDR of 1%. Then, we will ask NManout2 to output for each SNEP the functional annotations (as provided in the file Data/BY_S288c/Features/features.gff) falling symmetrically within a 5kb window around the SNEP. Finally, we will ask also to get for each SNEP the 5 nucleosomes before and the 5 nucleosomes after, in order to investigate if the SNEP is isolated or not. This multiple query can be done as follows:

$NManout2 -i S5/NManova -d HucH3K14_norm.db -o S5/NManout \
         -f Data/BY_S288c/Features/features.gff \
         -w 2500 -p 8.44e-4 -b 5

The command generates 3 files whith the stem name NManout in the folder S5:

• NManout_snep_table.txt is the file providing the result of the multiple query,
• NManout_snep.gff is a file in GFF format (version 3) listing all the SNEPs called at the chosen p-value cut-off. This file can then be used as input to your favorite genome browser.
• NManout_snep_tiling.txt is a file that will be useful to do some extra analysis on the SNEP distribution w.r.t to functional annotation using either NMt2feat or NMt2featcis.