robustness_pipeline issueshttps://gitlab.inria.fr/ml_genetics/private/robustness_pipeline/-/issues2020-03-17T14:44:49+01:00https://gitlab.inria.fr/ml_genetics/private/robustness_pipeline/-/issues/15mapdamage issue2020-03-17T14:44:49+01:00Mathieu MICHELmapdamage issueWhich file pass to mapdamge ? dnacomp / C to T frequency ?
Do we need to keep the reference sequence ?
For the moment just use Briggs model (4 values) and find out what the values areWhich file pass to mapdamge ? dnacomp / C to T frequency ?
Do we need to keep the reference sequence ?
For the moment just use Briggs model (4 values) and find out what the values areMathieu MICHELMathieu MICHELhttps://gitlab.inria.fr/ml_genetics/private/robustness_pipeline/-/issues/14mapping aDNA options2020-03-04T17:25:06+01:00JAY Floramapping aDNA optionshttps://gitlab.inria.fr/ml_genetics/private/robustness_pipeline/blob/master/bash_script/individual_Garga_and_postTreatment.sh#L40
should we add specific options for aDNA mapping ?
by default only the first (eg 10) bp of the reads are ch...https://gitlab.inria.fr/ml_genetics/private/robustness_pipeline/blob/master/bash_script/individual_Garga_and_postTreatment.sh#L40
should we add specific options for aDNA mapping ?
by default only the first (eg 10) bp of the reads are checked for errors, to fasten the proceess but this is not good for aDNA as there is more damage at the read's extremities
-> check the full read instead and allow for a certain percentage of mismatch (eg 10% of the read)https://gitlab.inria.fr/ml_genetics/private/robustness_pipeline/-/issues/12optimize mapping + conversion to bam into a single command line2020-03-04T16:14:51+01:00JAY Floraoptimize mapping + conversion to bam into a single command linehttps://gitlab.inria.fr/ml_genetics/private/robustness_pipeline/blob/master/bash_script/individual_Garga_and_postTreatment.sh#L41
(ask Maria)https://gitlab.inria.fr/ml_genetics/private/robustness_pipeline/blob/master/bash_script/individual_Garga_and_postTreatment.sh#L41
(ask Maria)Mathieu MICHELMathieu MICHELhttps://gitlab.inria.fr/ml_genetics/private/robustness_pipeline/-/issues/4change nucleotide at derived position: combersome ?2020-02-25T16:11:29+01:00JAY Florachange nucleotide at derived position: combersome ?https://gitlab.inria.fr/ml_genetics/private/robustness_pipeline/blob/master/npz_to_fasta.py#L39-41
c'est qqchose que vous aviez fait avec Agnes qui me parait compliqué et potentiellement peu efficace ?
pourquoi sauver toutes les choses ...https://gitlab.inria.fr/ml_genetics/private/robustness_pipeline/blob/master/npz_to_fasta.py#L39-41
c'est qqchose que vous aviez fait avec Agnes qui me parait compliqué et potentiellement peu efficace ?
pourquoi sauver toutes les choses à changer dans un dictionnaire (fonction searchSNP) puis faire les changements un à un? je verrais plutot un truc du genre
```
def apply_mutations(pos, nuc, aln, snp_col, comp):
derived = random draw from comp[nuc]
aln[snp_col==1, pos] = derived # assignation avec mask plus efficace qu'une boucle
define comp dictionnary
[apply_mutations(pos, seq[pos], aln_anc, simu[:,col], comp) for col,pos in enumerate(pos_simu)] # où pos_simu serait le vecteur des positions variantes conservé du read_ms() (présentement ce vecteur est overwritten)
```
a priori le changement est bien répercuté hors de la fonction et ca me semble plus straigtforwardJean CuryJean Cury