You may use both IUPAC codes and regular expression bracket expressions. For example "NTRGGTCAN" and "[ACGT]T[AG]GGTCA[ACGT]" are equivalent approximations of the example query shown on the right.

A motif is generated from the specified IUPAC sequence fragment by distributing the site counts equally over each possible residue. The pseudocount is added to each residue in proportion to the background. Thus the total count for a position is the site count plus the pseudocount.

For example suppose one of the IUPAC codes entered is 'R'. The IUPAC code 'R' means that both Adenine and Guanine are possible, so with a site count of 10, pseudocount of 1 and a uniform background; the probability of Adenine and Guanine would be:
while the probility of Cytosine and Thymine would be:

The background is used to determine how the pseudocount is distributed. While it is conventional to have the background sum to one these fields will be normalised so you don't have to be exact. You may even use counts instead.

A matrix is rectangular grid of numbers which shows the relative frequency a nucleotide will occur at a specific position. Two orientations are possible. For example:

Tomtom automatically detects the orientation but if for some reason the correct orientation isn't being detected it can be specified in the advanced section.

A matrix is column oriented if it has 4 rows and any number of columns. The original Tomtom web format was column oriented. A matrix is row oriented if it has 4 columns and any number of rows.

The site count is only used when a probability matrix is specified, otherwise it is derived from the matrix.

The pseudocount is added to each residue in proportion to the background. Thus the total count for a position is the site count plus the pseudocount.

By default only the first motif is searched but this can be changed in the advanced section. To search with more than one motif Tomtom must process your request as part of a queue and hence a contact email will be required. See more information on the meme format and the sample MEME DNA motifs.

The filter supports selecting motifs from the file by their name and/or their order in the file. Any numbers are assumed to refer to the position of the motif in the file. Any other entry is assumed to be a motif name (or ID). Motif names can not start with a dash and can only contain alphanumeric characters plus : _ . and -. The motifs should be separated by spaces.

An uploaded database is a normal MEME DNA motif file with lots of motifs. The uploaded database must be smaller than 5MB.

The email is used to notify you of your results but is not stored.

The job description will be included in the notification email you receive.

Motifs are compared column by column. In the following equations X is the set of probabilities from one column and Y the probabilities from the compared column. The subscript a refers to the letter from the nucleotide alphabet A.

Pearson correlation coefficient

Euclidean distance

Sandelin-Wasserman similarity

There are two ways of specifying the threshold.

E-value

The expected number of false positives. This may be any positive number.

q-value

Minimum False Discovery Rate at which a result is declared significant. This is a number between zero and one non-inclusive.

Complete scoring takes into account unaligned columns in the calculation of scores. It is recommended that you leave this feature enabled unless you are trying to recreate old results.

The background is used to determine how the pseudocount is distributed. While it is conventional to have the background sum to one these fields will be normalised so you don't have to be exact. You may even use counts instead.

Tomtom can search databases installed on the server or search a user uploaded database in MEME DNA motif format by selecting the "Upload Your Own Database" option. Uploaded databases are searched in a queue and so require an email address. For more information browse the supported databases.