@@ -5,7 +5,7 @@ layout: default
55# Further Examples
66
77Continuing on from the examples in [ Getting Started with CountESS] ( ../getting-started/ )
8- let's look at a more realistic experiment.
8+ let's look at some more realistic experiments:
99
1010## Example 5: Deep Mutational Scan of BRCA1
1111
@@ -107,28 +107,66 @@ before calculating scores as above:
107107
108108[ ![ Screenshot 12] ( img/s5_12.jpg )] ( img/s5_12.png )
109109
110- [ ![ Screenshot 13] ( img/s5_13.jpg )] ( img/s5_13.png )
110+ <!-- [](img/s5_13.png)-->
111+
112+ This will let us calculate scores and sigmas per
113+ DNA variant instead of per barcode:
111114
112115[ ![ Screenshot 14] ( img/s5_14.jpg )] ( img/s5_14.png )
113116
114- [ ![ Screenshot 16] ( img/s5_16.jpg )] ( img/s5_16.png )
117+ Scores can then be combined just like the "per barcode"
118+ example:
119+ <!-- [](img/s5_16.png)-->
115120
116121[ ![ Screenshot 17] ( img/s5_17.jpg )] ( img/s5_17.png )
117122
118123### Scoring by Protein
119124
120- ... Or to collate by protein variant:
125+ ... Or we can collate by protein variant:
121126
122127[ ![ Screenshot 21] ( img/s5_21.jpg )] ( img/s5_21.png )
123128
124- [ ![ Screenshot 22] ( img/s5_22.jpg )] ( img/s5_22.png )
129+ <!-- [](img/s5_22.png)
125130
126131[](img/s5_15.png)
127132
128- [ ![ Screenshot 18] ( img/s5_18.jpg )] ( img/s5_18.png )
133+ [](img/s5_18.png)-->
129134
130135[ ![ Screenshot 19] ( img/s5_19.jpg )] ( img/s5_19.png )
131136
132137
138+ ### Other Options
139+
140+ The above example uses two replicates and three time points.
141+ If only two timepoints are available, the scoring plugin falls
142+ back on the "log ratio" version of the scoring formula, also
143+ as per Enrich2.
144+
145+ If there's no suitable "wild type indicator" column, the
146+ scoring plugin can use the total of all variants instead, also
147+ as per Enrich2.
148+
149+ ## Example 6: VAMPseq
150+
151+ [ VAMP-seq] ( https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5980760/ )
152+ sorts cells into bins and uses a weighted sum of the frequencies of
153+ variants in each bin to calculate a score for each variant.
154+
155+ In this example we have three replicates each of which has counts in
156+ four bins.
157+
158+ The first step is to load and collate these files just like in
159+ [ Example 3] ( ../getting-started/#example-3 ) :
160+
161+ [ ![ Screenshot 1] ( img/s6_1.jpg )] ( img/s6_1.png )
162+
163+ Then we can use the [ VAMPseq Plugin] ( ../included-plugins/#vampseq ) to calculate scores
164+ for each variant:
165+
166+ [ ![ Screenshot 2] ( img/s6_2.jpg )] ( img/s6_2.png )
133167
168+ Once that's done, pivot on replicate and use a simple formula to combine scores
169+ into an average score and an estimated standard deviation:
134170
171+ [ ![ Screenshot 3] ( img/s6_3.jpg )] ( img/s6_3.png )
172+ [ ![ Screenshot 4] ( img/s6_4.jpg )] ( img/s6_4.png )
0 commit comments