More than one TF per volcano dot

[RadPa]

Hi, I have a few questions regarding the binding scores and the volcano plot created by TOBIAS.

1. Can more than one TF have identical binding scores? I observed so in my results.txt.
2. I understand that each dot in the volcano plot is a TF motif; when I plotted from results.txt (after the percentile cutoff), a dot has more than one TF (2 to 10).
3. Plot created by BINDetect has only one TF per dot. Snippet from results.txt.

   B-H1_MA0168.1   B-H1    MA0168.1    C_unc-4 59759   0.12574 3399    0.12386 3135    0.05502 3.29476E-111
   B-H2_MA0169.1   B-H2    MA0169.1    C_unc-4 59759   0.12574 3399    0.12386 3135    0.05502 3.29476E-111
   C15_MA0170.1    C15 MA0170.1    C_unc-4 59759   0.12574 3399    0.12386 3135    0.05502 3.29476E-111
   lms_MA0175.1    lms MA0175.1    C_unc-4 59759   0.12574 3399    0.12386 3135    0.05502 3.29476E-111
   NK7.1_MA0196.1  NK7.1   MA0196.1    C_unc-4 59759   0.12574 3399    0.12386 3135    0.05502 3.29476E-111
   abd-A_MA0206.1  abd-A   MA0206.1    C_lbl   67779   0.12115 3287    0.11933 2961    0.05288 2.58402E-118
   achi_MA0207.1   achi    MA0207.1    C_vis   64161   0.15151 8429    0.15211 8427    -0.00909    6.42775E-41
   al_MA0208.1 al  MA0208.1    C_lbl   67779   0.12115 3287    0.11933 2961    0.05288 2.58402E-118

   I am not sure if I am doing something wrong.

Thank you Radhika

[msbentsen]

Hi Radhika,

If two or more motifs are similar (and thus the same binding sites are found), then the scores will be identical in the results. Is this the case here? I see that the motifs have different ID's, but if the sites in output/<TF>/beds/<TF>_all.bed are identical, that explains it.

If that is not the case, I would need a little more information about your run e.g. TOBIAS version, which type of data you used, commands etc. to find out what is going on. Thanks!

[RadPa]

Thank you for the explanation. I checked the bed files and the sites are identical; it makes sense. I would like to know how the BINDetect labels only one TF as significant in the volcano plot. I apologize for the delayed response.

Thank you Radhika

[msbentsen]

Hi Radhika,

BINDetect labels the top 5% TFs in each direction in the volcano plot, so I guess it can happen that two identical TFs lie right at the 5% border (so one in, one out). Is that what you meant? In that case, you can count all the TFs at this position as "significant".

BR Mette

[RadPa]

Hi Mette, I apologize for the delay. Yes, that's what I was trying to ask. Thank you Radhika