Alignment data
Contents |
[edit] 1 Iterations
Iterations done for each class until no value in the scaled log odds matrix deviated more than 1 from the previous iteration's scaled log odds matrix.
| Class | Disordered Iterations | Ordered Iterations |
|---|---|---|
| 85% | 3 | 3 |
| 60% banded | 4 | 3 |
| 40% banded | 3 | 3 |
[edit] 2 Number of alignments
Both ordered and disordered having 213 families is a coincidence.
Disordered has fewer alignments per sequence than ordered. Disordered number of families decreases in lower banding, not increases. Banding increases included ordered families, however.
Note from previous: banded is not simply the unbanded minus the upper echelons, because the unbanded matrix classes will include things like 85% identity with over 1 gap. Banding can decrease the number of families present, presumably ones containing only similar sequences. Banding increases the percent mutation rate. Banding increases the sum off the nondiagonal. Banding closes the gap between the ordered and disordered mutation rate.
| Num families | Num seqs | Num alignments | Num substitutions | |
|---|---|---|---|---|
| Ordered 85% | 213 | 3,408 | 38,117 | 6,976,696 |
| Ordered 60% banded | 224 | 27,316 | 3,327,744 | 590,311,337 |
| Ordered 40% banded | 242 | 52,527 | 5,949,331 | 936,880,297 |
| Disordered 85% | 213 | 3,127 | 29,259 | 2,284,355 |
| Disordered 60% banded | 207 | 18,883 | 662,599 | 50,474,512 |
| Disordered 40% banded | 182 | 31,361 | 1,738,606 | 147,793,757 |
[edit] 2.1 Gap cutoff effects
Shows how many alignments are cut off by maximum gap limits.
| Total alignments | Gap cutoffs | Included alignments | Percent cutoff | |
|---|---|---|---|---|
| Ordered 85% | 68,397 | 0 | 41,551 | 39.250 |
| Disordered 85% | 76,840 | 0 | 29,259 | 61.922 |
| Ordered 60% banded | 3,483,138 | 4 | 3,327,744 | 4.461 |
| Disordered 60% banded | 750,526 | 4 | 662,599 | 11.715 |
| Ordered 40% banded | 8,548,724 | 4 | 5,949,331 | 30.407 |
| Disordered 40% banded | 2,417,993 | 4 | 1,738,606 | 28.097 |
[edit] 3 Percent identity quantiles
Rejoicing: ordered and disordered have very similar PIs now. Gap cutoffs do not seem to reduce the average PI by much, either, also good.
[edit] 3.1 With gap cutoffs
| Gap cutoff | 0% | 25% | 50% | 75% | 100% | Average | |
|---|---|---|---|---|---|---|---|
| Ordered 85% | 0 | 85.000 | 88.186 | 92.063 | 96.429 | 99.890 | 92.285 |
| Disordered 85% | 0 | 85.000 | 87.879 | 91.304 | 94.737 | 99.932 | 91.521 |
| Ordered 60% banded | 4 | 60.000 | 63.953 | 68.468 | 74.107 | 85.000 | 69.365 |
| Disordered 60% banded | 4 | 60.000 | 63.636 | 68.571 | 75.510 | 85.000 | 69.826 |
| Ordered 40% banded | 4 | 40.000 | 45.033 | 49.701 | 54.651 | 60.000 | 49.855 |
| Disordered 40% banded | 4 | 40.000 | 43.704 | 48.246 | 53.731 | 60.000 | 48.841 |
[edit] 3.2 Without gap cutoffs
| 0% | 25% | 50% | 75% | 100% | Average | |
|---|---|---|---|---|---|---|
| Ordered 85% | 85.000 | 87.143 | 89.710 | 94.898 | 99.890 | 90.977 |
| Disordered 85% | 85.000 | 87.273 | 89.815 | 93.548 | 99.932 | 90.510 |
| Ordered 60% banded | 60.000 | 63.799 | 68.158 | 73.889 | 85.000 | 69.142 |
| Disordered 60% banded | 60.000 | 63.415 | 68.085 | 75.000 | 85.000 | 69.521 |
| Ordered 40% banded | 40.000 | 43.946 | 48.168 | 53.285 | 60.000 | 48.771 |
| Disordered 40% banded | 40.000 | 43.704 | 48.246 | 53.731 | 60.000 | 47.822 |
[edit] 4 Percent identity histograms
[edit] 4.1 85%
[edit] 4.2 60% banded
[edit] 4.3 40% banded
[edit] 5 Family contributions
Note: families with no alignments in a class are excluded from these charts. In these boxplots, the midline is the median and the box extends from the first to the third quartile. The whiskers extend to the most extreme data point no more than interquartile range from the box. (?)
Observations: ordered families have higher median family contributions. That means fluctuations in families should bother them less.
[edit] 6 Sequences
[edit] 7 Alignments
[edit] 8 Substitutions
