Skip to content

Latest commit

 

History

History
148 lines (127 loc) · 6.08 KB

compute.md

File metadata and controls

148 lines (127 loc) · 6.08 KB

Goalign: toolkit and api for alignment manipulation

Commands

compute

This command implements different computations:

  1. goalign compute distance : Computes a distance matrix from an input DNA alignment, with different evolutionary models. In the case of ambigous nucleotides (IUPAC), one mutation is counted if characters are not compatible (ex: R vs. Y). Possible models are:
    • pdist
      • rawdist : Raw distance (like pdist, without normalization by length)
    • jc : Juke-Cantor
    • k2p : Kimura 2 Parameters
    • f81 : Felsenstein 81
    • f84 : Felsenstein 84
    • tn93 : Tamura and Nei 1993 If distance is pdist (nucleotides), then giving the option --rm-ambiguous will not take into account ambiguous positions that compatible, for length normalization. For example if --rm-ambiguous is given, then R vs. Y will be taken into account because there is a difference. And N vs. A won't be taken into account in total length because we are not sure whether they are identical. In case of a nucleotidic alignment, it is possible to specify sequence ranges to compare. For example, goalign compute distance -m pdist -i align.ph -p --range1 0:9 --range2 10:19 will compute distance only between sequences [0 to 9] and sequences [10 to 19]. Output matrix will be formatted the same way as usual, except that it will be made of 0 except for the comparisons 0 vs. 10; 0 .vs 11; ...; 9 vs. 19.
  2. goalign compute entropy: Computes the entropy of each sites of the input alignment or the average entropy of all sites (-a option).
  3. goalign compute pssm: Computes and prints a Position specific scoring matrix. Different kind of matrices may be computed, depending on -n option:
    • -n 0 : None, means raw counts
    • -n 1 : By column frequency, i.e. frequency of nt/aa per site/column
    • -n 2 : By column frequency compared to alignment frequency: same as -n 1, but divides by frequency of the nt/aa in the whole alignment
    • -n 3 : By column frequency compared to uniform frequency: same as -n 1, but divides by uniform frequency of the nt/aa (1/4 for nt, 1/20 for aa)
    • -n 4 : Normalization "Logo". Option -c allows to add pseudo counts before normalization, and option -l log2 transforms the values.

Usage

  • General command
Usage:
  goalign compute [command]

Available Commands:
  distance    Compute distance matrix from an input alignment
  entropy     Computes entropy of a given alignment
  pssm        Computes and prints a Position specific scoring matrix

Flags:
  -h, --help   help for compute

Global Flags:
  -i, --align string   Alignment input file (default "stdin")
  -p, --phylip         Alignment is in phylip? False=Fasta
  --input-strict       Strict phylip input format (only used with -p)
  • distance command
goalign compute distance [flags]

Flags:
  -a, --average         Compute only the average distance between all pairs of sequences
  -m, --model string    Model for distance computation (default "k2p")
  -o, --output string   Distance matrix output file (default "stdout")
  -r, --rm-gaps         Do not take into account positions containing >=1 gaps

Global Flags:
  -i, --align string   Alignment input file (default "stdin")
  -p, --phylip         Alignment is in phylip? False=Fasta
  --input-strict       Strict phylip input format (only used with -p)
  • pssm command
Usage:
  goalign compute pssm [flags]

Flags:
  -l, --log                   (normalized) Values in log2
  -n, --normalization int     Counts normalization
  -c, --pseudo-counts float   Value added to (normalized) counts

Global Flags:
  -i, --align string   Alignment input file (default "stdin")
  -p, --phylip         Alignment is in phylip? False=Fasta

Examples

  • Generating a random tree with 5 tips (Gotree), simulating an alignment from this tree (seq-gen, and computing a distance matrix (model f81) from this alignment:
gotree generate yuletree -l 5 --seed 1 -o true_tree.nw
seq-gen -op -mGTR -l500 -z 2 -n 1 true_tree.nw > alignment.phy
goalign compute distance -i alignment.phy -m f81 -o dist.txt -p -t 10

Should give the following distance matrix:

5
Tip4    0.000000000000  0.174911845895  0.192803956978  0.232646053483  0.235379041630
Tip0    0.174911845895  0.000000000000  0.082364641962  0.128396525775  0.142776083476
Tip3    0.192803956978  0.082364641962  0.000000000000  0.071285264523  0.086842665158
Tip2    0.232646053483  0.128396525775  0.071285264523  0.000000000000  0.111961817720
Tip1    0.235379041630  0.142776083476  0.086842665158  0.111961817720  0.000000000000
  • Generating a random tree with 100 tips (Gotree), simulating an alignment from this tree (seq-gen, and computing entropy of each site of this alignment:
gotree generate yuletree -l 200 --seed 1 -o true_tree.nw
seq-gen -op -mGTR -l10 -z 2 -n 1 true_tree.nw > alignment.phy
goalign compute entropy  -i alignment.phy -p

Should give the following matrix:

Alignment  Site  Entropy
0          0     0.616
0          1     0.858
0          2     0.806
0          3     0.621
0          4     0.621
0          5     0.758
0          6     0.813
0          7     0.780
0          8     0.762
0          9     0.816
  • Generating a random tree with 100 tips (Gotree), simulating an alignment from this tree (seq-gen), and computing a logo from this alignment:
gotree generate yuletree -l 200 --seed 1 -o true_tree.nw
seq-gen -op -mGTR -l10 -z 2 -n 1 true_tree.nw > alignment.phy
goalign compute pssm -n 4 -i alignment.phy -p -c 0.0001

Should give the following matrix:

    A      C      G      T
1   0.089  0.901  0.122  0.000
2   0.541  0.038  0.038  0.145
3   0.628  0.033  0.117  0.059
4   0.088  0.022  0.077  0.916
5   0.022  0.077  0.916  0.088
6   0.118  0.063  0.027  0.698
7   0.066  0.628  0.058  0.074
8   0.096  0.674  0.061  0.044
9   0.054  0.703  0.054  0.090
10  0.041  0.576  0.189  0.016