# Benchmark 3 v1: oriented contig and region order results analysis

Only the non-perfectly successful instances are analysed.


## Agathis_dammara

### Sufficient fail causes

* $1_{mult} = 2$, while it should be $3$.
    * Indeed, $1_{cov} = 223$ and $19_{cov} = 121$, which is the starter

### Other observations

* contig $0$, contig $1$, contig $2$ have no protein-DNA alignment length


## Begonia_pulchrifolia

### Sufficient fail causes

* The starter $0_{mult} = 1$ while contig $0$ should belong to inverted repeat.


## Carpodetus_serratus

### Sufficient fail causes

* Inversion in a single copy of ($6_f$) $10_f 12_r 1_f$ ($11_f$) that becomes ($6_f$) $1_r 12_f 10_r$ ($11_f$)
    * Due to links $(6_f, 10_f)$, $(6_f, 1_r)$, $(1_f, 11_f)$ and $(10_r, 11_f)$

### Explanation

The links $(11_r, 1_r)$ and $(10_r, 6_r)$ hide a little inverted repeat (74-75 bp), that Quast reveals:
* $ir$
    * reference: 5 581 – 5 654 (74) bp
    * contig: 77 384 – 77 457 (74) bp
* $\overline{ir}$
    * reference: 67 134 – 67 208 (75) bp
    * contig: 77 457 – 77 383 (75) bp


## Cucumix_hystrix

### Sufficient fail causes

* IR are not equals, one IR has a SC of length $1$ in it
    * Although $7_{mult} = 1$ (expected), it is not taken otherwise it would decrease the value of the objective function.


## Jasminum_tortuosum

### Sufficient fail causes

* Missing link $(6_f, 4_f)$
    * In link set there is $(6_f, 4_r)$ (obtained as a result) but no $(6_f, 4_f)$ (expected).
    * Surprisingly, there are $(9_f, 4_f)$ and $(9_f, 4_r)$


## Lamprocapnos_spectabilis

### Sufficient fail causes

* The starter $4_{mult} = 1$, while it should be $2$ as contig $4$ belongs to an inverted repeat.


## Lathyrus_pubescens

### Sufficient fail causes

* Missing vertex contig $12_f$ in the single-copy between $15_r 12_f 13_r$: it may have participated in the path but because its $12_{prob} = 0$, it would not increase the value of the objective function.


## Lophocereus_schottii

### Sufficient fail causes

* Inversion of ($10_f$) $2_f 1_r 3_r$ ($5_f$) due to links $(10_f, 3_f)$, $(10_f, 2_f)$, $(2_r, 5_f)$ and $(3_r, 5_f)$.

### Explanation

The links $(10_f, 2_f)$ and $(3_r, 5_f)$ hide a little inverted repeat (171 bp):
* By aligning the reference at the positions 40kb-44kb to itself at positions 73kb-76kb, blast finds a perfect identity score at 42547-42717 aligned with 74795-74625 (plus/minus)
* Note that $(10_f, 2_f)$ and $(10_f, 3_f)$ are highly similar (some SNPs)


## Pelargonium_nanum

### Sufficient fail causes

* Get in IR $2_r$ instead of $2_r 10_f 9_f 2_r$: $2_{mult} = 3$ while it should be $= 4$.


## Podocarpus_totara

### Sufficient fail causes

* Link $(6_f, 11_r)$ has been taken instead of $(6_f, 7_r)$ ; so link $(2_f, 7_r)$ has been taken instead of $(2_f, 10_f)$


## Triosteum_pinnatifidum

### Sufficient fail causes

* All good except that contig $7$ is not chosen in the single-copy as $7_{prob} = 0$