New multi-sample toy data set

[abremges]

For the second mode of Opal (cross-sample comparisons to check how well tools capture e.g. changes in alpha diversity), we need to simulate a small and controlled data set, generate the gold standard, and run different profilers on it.

@fernandomeyer and I will think about it and discuss the experimental setup (probably involving @AlphaSquad, too, if CAMISIM can do the job).

[alicemchardy]

We can use the medium complexity data set and results for this - it is diff abundance. Also the time series high complexity could be analysed that way.

Best, Alice

Sent from my phone

Am 15.11.2017 um 12:13 schrieb Andreas Bremges notifications@github.com<mailto:notifications@github.com>:

For the second mode of Opal (cross-sample comparisons to check how well tools capture e.g. changes in alpha diversity), we need to simulate a small and controlled data set, generate the gold standard, and run different profilers on it.

@fernandomeyerhttps://github.com/fernandomeyer and I will think about it and discuss the experimental setup (probably involving @AlphaSquadhttps://github.com/alphasquad, too, if CAMISIM can do the job).

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/CAMI-challenge/OPAL/issues/20, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ACAH4QurRbA8YJqGR9AiaYcxql40hl_Rks5s2sdVgaJpZM4Qew86.

---

Helmholtz-Zentrum für Infektionsforschung GmbH | Inhoffenstraße 7 | 38124 Braunschweig | www.helmholtz-hzi.de

Vorsitzende des Aufsichtsrates: MinDir’in Bärbel Brumme-Bothe, Bundesministerium für Bildung und Forschung Stellvertreter: MinDirig Rüdiger Eichel, Niedersächsisches Ministerium für Wissenschaft und Kultur Geschäftsführung: Prof. Dr. Dirk Heinz; Silke Tannapfel Gesellschaft mit beschränkter Haftung (GmbH) Sitz der Gesellschaft: Braunschweig Handelsregister: Amtsgericht Braunschweig, HRB 477

[abremges]

But CAMI's MC and HC data sets have the same alpha diversity (same species occur with variable abundances). I thought one aspect was to assess how well tools detect varying community complexity!?

[alicemchardy]

Hmm true.. though they might still fail on that on these datasets, as sometimrs abundances are very low..

Re: measures of complexity, got some input from Ruben, who likes Shannon ( for abundance, though Davids examples made me even more worried), some tree-based index (childs? Or faith? Will check) and simply counting taxa for comparing abundances at a certain sampling depth and rank, maybe rarefaction curves would be a good/better idea?

Best, Alice

Sent from my phone

Am 15.11.2017 um 18:02 schrieb Andreas Bremges notifications@github.com<mailto:notifications@github.com>:

But CAMI's MC and HC data sets have the same alpha diversity (same species occur with variable abundances). I thought one aspect was to assess how well tools detect varying community complexity!?

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/CAMI-challenge/OPAL/issues/20#issuecomment-344658910, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ACAH4YfFQSeHM_mGvSrwWVYgxkAY9B1nks5s2xkwgaJpZM4Qew86.

---

Helmholtz-Zentrum für Infektionsforschung GmbH | Inhoffenstraße 7 | 38124 Braunschweig | www.helmholtz-hzi.de

Vorsitzende des Aufsichtsrates: MinDir’in Bärbel Brumme-Bothe, Bundesministerium für Bildung und Forschung Stellvertreter: MinDirig Rüdiger Eichel, Niedersächsisches Ministerium für Wissenschaft und Kultur Geschäftsführung: Prof. Dr. Dirk Heinz; Silke Tannapfel Gesellschaft mit beschränkter Haftung (GmbH) Sitz der Gesellschaft: Braunschweig Handelsregister: Amtsgericht Braunschweig, HRB 477

[sjanssen2]

I always look at "observed_otus" (= raw number of different species regardless of their abundance), "shannon" and "Faith Phylogenetic Diversity" (which requires a phylogenetic tree and assess covered branch lengths). I like the latter the most, but do we have a phylogeny for organisms in CAMI data? You might want to take a look at https://docs.qiime2.org/2017.10/plugins/available/diversity/alpha-rarefaction/ for rarefaction curves

[alicemchardy]

These are the same ones, perfect! We habe the taxonomy, for Faith.

Best, Alice

Sent from my phone

Am 15.11.2017 um 20:10 schrieb Stefan Janssen notifications@github.com<mailto:notifications@github.com>:

I always look at "observed_otus" (= raw number of different species regardless of their abundance), "shannon" and "Faith Phylogenetic Diversity" (which requires a phylogenetic tree and assess covered branch lengths). I like the latter the most, but do we have a phylogeny for organisms in CAMI data? You might want to take a look at https://docs.qiime2.org/2017.10/plugins/available/diversity/alpha-rarefaction/ for rarefaction curves

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/CAMI-challenge/OPAL/issues/20#issuecomment-344696417, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ACAH4ZpcAk7EzJTirBKIZ6SdVWLPY7rWks5s2zcbgaJpZM4Qew86.

---

Helmholtz-Zentrum für Infektionsforschung GmbH | Inhoffenstraße 7 | 38124 Braunschweig | www.helmholtz-hzi.de

Vorsitzende des Aufsichtsrates: MinDir’in Bärbel Brumme-Bothe, Bundesministerium für Bildung und Forschung Stellvertreter: MinDirig Rüdiger Eichel, Niedersächsisches Ministerium für Wissenschaft und Kultur Geschäftsführung: Prof. Dr. Dirk Heinz; Silke Tannapfel Gesellschaft mit beschränkter Haftung (GmbH) Sitz der Gesellschaft: Braunschweig Handelsregister: Amtsgericht Braunschweig, HRB 477

[alicemchardy]

Correction: for CAMI we so far only have predictions for the pooled data sets, not individual CAMI samples.. So at least we have to rerun indivdiual tools for individual samples.

[fernandomeyer]

Cross-sample comparisons have been done using CAMI I high and CAMI II mouse gut data. Alpha/beta diversities can now be measured with Shannon, rarefaction curves...