|
|||||||||||||||
 |
|
||||||||||||||
|
|||||||||||||||
| |||||||||||||||
|
Score sequence data with a PSSMDescriptionPosisiton specific scoring matices (PSSM) are more statistially motivated sequence motif models that provide higher sensitivity and specificity than regular expressions or ad hoc approaches (see project 08).
Input and output The program is given a fasta file with DNA or amino acid sequences, a PSSM and a score threshold. You can find weight matices that describe the DNA binding preference for transcription factors in the database TRANSFAC. You will have to register with TRANSFAC to gain access but it is free. The units for these weights are often unknown and as such should be interpreted as un-normalized frequencies. The output should be the sites that score above the provided threshold. Site scores are calculated as the sum of PSSM weights selected at each position.
Statistical analysis Frequencies can be interptreted as letter (nucleotide or amino acid) probabilities for a particular letter at a particular position in the matrix. The letter background probabilities, i.e. probability of a letter not within the pattern, are aften called prior probabilities (priors), i.e. what we would expect before we knew about the motif. The priors will depend on the organism and how the background is defined. In E.coli, for example, the genomic nucleotide priors are almost equiprobable (p_A:0.25, p_C:0.25, p_G:0.25, p_T:0.25) but in yeast they are closer to p_A:0.31, p_C:0.19, p_G:0.19, p_T:0.31 (note that p_A==p_T and p_C==p_G). TRANSFAC example M00204 for GCN4 (yeast, Saccharomyces cerevisiae transcription factor). PO A C G T 01 2.69 2.46 1.74 0.81 N 02 5.18 0.00 2.52 0.00 A 03 4.30 0.00 2.43 0.97 R 04 0.00 0.00 0.00 7.70 T 05 0.00 0.00 7.70 0.00 G 06 7.70 0.00 0.00 0.00 A 07 0.00 6.01 0.00 1.68 C 08 0.00 0.00 0.00 7.70 T 09 0.00 7.70 0.00 0.00 C 10 3.40 1.65 0.00 1.88 WYou may also wish to generate an alignment counts matrix from sequence examples like these, GGCTTACAAAC TTATTACTAAT TGCTTATAAAT CGCTGACTAAT CGCTGACTAAT TGCTTACTAGT TGCTTACTAGT TGATTAATAAT TGATTAATAAT TGCTGACTAAA TTTTTACTAAT TTTTTACTAATBut as there are potentially zeros in these matrices (not very likely reflecting reality since we typically have a small sample and there are only 4 letters) we need to add pseudo-counts to avoid observed frequencies (o_A, o_C, o_G, o_T) being zero.
Probability matrix of the above with 1 added pseudo-count, q=1
where, for position i, PO A C G T 01 0.338 0.311 0.229 0.122 02 0.624 0.029 0.318 0.029 03 0.523 0.029 0.308 0.140 04 0.029 0.029 0.029 0.914 05 0.029 0.029 0.914 0.029 06 0.914 0.029 0.029 0.029 07 0.029 0.720 0.029 0.222 08 0.029 0.029 0.029 0.914 09 0.029 0.914 0.029 0.029 10 0.460 0.240 0.032 0.269A probability matrix is then converted to a log-likelihood matrix h_iA = log2(f_iA/p_A) h_iC = log2(f_iC/p_C) ... and matrix H looks like this, PO A C G T 01 0.435 0.317 -0.128 -1.037 02 1.320 -3.121 0.349 -3.121 03 1.065 -3.121 0.301 -0.834 04 -3.121 -3.121 -3.121 1.870 05 -3.121 -3.121 1.870 -3.121 06 1.870 -3.121 -3.121 -3.121 07 -3.119 1.527 -3.119 -0.171 08 -3.121 -3.121 -3.121 1.870 09 -3.121 1.870 -3.121 -3.121 10 0.881 -0.061 -2.987 0.104 The fasta file can contain more than one sequence/entry.
|
| Follow CBS on Twitter |
