Source for number of variants and samples on main page

[wm75]

Currently, the main page states this under Results for Genomics:

These lists are updated daily. There are 397 sites showing intra-host variation across 33 samples (with frequencies between 5% and 95%). Twenty nine samples have fixed differences at 39 sites from the published reference.

This leaves two questions:

1. when have these numbers been updated last (based on which samples)?
2. when is a variant considered a fixed difference?

When I analyzed https://covid19.galaxyproject.org/genomics/4-Variation/variant_list.tsv today, I got this for the filter condition 0.95 >= float(af) >= 0.05:
Samples with variants: 378
Total number of variants observed: 260
Number of sites observed to carry variants: 259

For the fixed differences I tried float(af) == 1.0 giving:
Samples with variants: 55
Total number of variants observed: 27
Number of sites observed to carry variants: 27

and float(af) > 0.95 resulting in:
Samples with variants: 378
Total number of variants observed: 260
Number of sites observed to carry variants: 259

[mvdbeek]

1. when have these numbers been updated last (based on which samples)?

1d272c3
27 days ago. We need to template this, it's clearly way more than 397

When I analyzed https://covid19.galaxyproject.org/genomics/4-Variation/variant_list.tsv today, I got this for the filter condition 0.95 >= float(af) >= 0.05:
Samples with variants: 378
Total number of variants observed: 260
Number of sites observed to carry variants: 259

How'd you do this? I just downloaded the file and there are 1410 unique positions that fall into this criterium:

import pandas as pd

df = pd.read_csv('https://covid19.galaxyproject.org/genomics/4-Variation/variant_list.tsv', sep='\t')
intrahost = df[(df['AF'] > 0.05) & (df['AF'] < 0.95) ]
len(intrahost.POS.unique())

returns 1410. More if you also count by variant.

[wm75]

Ah, I pasted the same result snippet twice above, sorry!
The first result should have been:
Samples with variants: 334
Total number of variants observed: 1466
Number of sites observed to carry variants: 1424

[wm75]

... and if I go with the filter 0.95 > float(af) > 0.05 instead of my initial >=:
Samples with variants: 333
Total number of variants observed: 1452
Number of sites observed to carry variants: 1410

which seems to be what you obtained, right?

[wm75]

I used plain Python for this:

def calc_latest_stats(fun):
    with open('va_current.tsv') as i:
        hdr = next(i)
        samples = set()
        variants = set()
        sites = set()
        for line in i:
            s, c, p, r, a, d, af, *others = line.split('\t')
            if fun(s, c, p, a, af) is True:
                samples.add(s)
                variants.add((c, p, a))
                sites.add((c, p))

    print('Samples with variants:', len(samples))
    print('Total number of variants observed:', len(variants))
    print('Number of sites observed to carry variants:', len(sites))

with, e.g.,:
fun = lambda s, c, p, a, af: True if 0.95 > float(af) > 0.05 else False