search

stat-lab / EvalSVcallers

Evaluate the performances (precision and recall) of structural variation (SV) callers
32 stars 13 forks source link

Recall Values for INS calling via MELT #14

Open Mkddb opened 3 years ago

Mkddb commented 3 years ago

Hello Team,

I have a couple of queries, i need help with.

1) Can you help me understanding why the "True positives" are indicated in Table S9, in place of "Recall values". 2) Any specific reason to Denote "Provisional recall values" in "Table 1" and what does it signifies about the data over "Recall values". 3) Also, is it possible to calculate similar "Provisional recall values" for other variant classes such as Deletion, Duplication, Inversion, Insertion. If yes, Any suggestion for that ?

stat-lab commented 3 years ago

The DGV database contains no MEI data and the reference MEIs (true MEIs) of NA12878 are unavailable. Thus, we used the estimated number of MEIs present in an individual genome, which had been reported in the previous study (Sudmant et al., Nature 2015), to calculate provisional TP calls and recall values. You may use the data reported in the more recent study (Collins et al., Nature 2020).

2020/11/10 15:32、Mkddb notifications@github.comのメール:

Hello Team,

I have a couple of queries, i need help with.

Can you help me understanding why the "True positives" are indicated in Table S9, in place of "Recall values". Any specific reason to Denote "Provisional recall values" in "Table 1" and what does it signifies about the data over "Recall values". Also, is it possible to calculate similar "Provisional recall values" for other variant classes such as Deletion, Duplication, Inversion, Insertion. If yes, Any suggestion for that ? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/stat-lab/EvalSVcallers/issues/14, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADIBV3HI3TS6L4A4PVEP6RDSPDM7RANCNFSM4TQIC27A.

Mkddb commented 3 years ago

Okay,

Thanks for the kind reply. I completely agree with it.

I have another small query. There's RSS values from minimum 2 Reads to 12 reads, but no such values provided for RSS at 11 reads (for all tools/SV classes). I was not able to find any reason behind that. Will you please shed some light on this. why RSS at 11 reads is not given ? also, why 12 reads were taken as upper limit and not beyond that ?