diff --git a/README.md b/README.md
index c1084d6..3bab491 100644
--- a/README.md
+++ b/README.md
@@ -2,30 +2,30 @@
# varVAMP
-[![License: GPL v3](https://img.shields.io/badge/License-GPLv3-blue.svg)](https://www.gnu.org/licenses/gpl-3.0)
+[![License: GPL v3](https://img.shields.io/github/license/jonas-fuchs/varvamp)](https://www.gnu.org/licenses/gpl-3.0)
For a lot of virus genera it is difficult to design pan-specific primers. varVAMP solves this, by introducing ambiguous characters into primers and minimizes mismatches at the 3' end. Primers might not work for some sequences of your input alignment but should recognize the large majority.
**varVAMP comes in three different flavors:**
-
+
-**SANGER** *(coming soon)*: varVAMP searches for the very best primers and reports back an amplicon which can be used for PCR-based screening approaches.
+**SANGER**: varVAMP searches for the very best primers and reports back non-overlapping amplicons which can be used for PCR-based screening approaches.
**TILED**: varVAMP uses a graph based approach to design overlapping amplicons that tile the entire viral genome. This designs amplicons that are suitable for Oxford Nanopore or Illumina based full-genome sequencing.
**QPCR** *(coming soon)*: varVAMP searches for small amplicons with an internal primer for the probe. It minimizes temperature differences between the primers.
-This program is currently being developed and in an alpha state. You are welcome to use this software. If you successfully design primers, drop me a mail. It might be possible to collaborate!
+This program is currently being developed and in an alpha state. You are welcome to use this software. If you successfully design primers, drop me a mail. It might be possible to collaborate! Ideas and suggestions are highly welcome.
# Documentation
-* [Installation](docs/installation.md)
-* [Preparing the data](docs/preparing_the_data.md)
-* [Usage](docs/usage.md)
-* [Output](docs/output.md)
-* [How it works](docs/how_varvamp_works.md)
-* [FAQ](docs/FAQ.md)
+* [Installation](https://github.com/jonas-fuchs/varVAMP/blob/master/docs/installation.md)
+* [Preparing the data](https://github.com/jonas-fuchs/varVAMP/blob/master/docs/preparing_the_data.md)
+* [Usage](https://github.com/jonas-fuchs/varVAMP/blob/master/docs/usage.md)
+* [Output](https://github.com/jonas-fuchs/varVAMP/blob/master/docs/output.md)
+* [How it works](https://github.com/jonas-fuchs/varVAMP/blob/master/docs/how_varvamp_works.md)
+* [FAQ](https://github.com/jonas-fuchs/varVAMP/blob/master/docs/FAQ.md)
---
diff --git a/docs/FAQ.md b/docs/FAQ.md
index 8ee18f9..c00dae1 100644
--- a/docs/FAQ.md
+++ b/docs/FAQ.md
@@ -4,18 +4,22 @@
Start with setting your optimal length of your amplicon, the max length that you can tolerate and the overlap that you want to achieve. If you want to allow ambiguous characters, set them to 2 as a start. Set the threshold to the mean sequence identity of your alignment. Run varvamp and then optimize the output.
-2. **How do I optimize the output?**
+2. **How do I optimize the output for TILED mode?**
It all depends on how many conserved regions varVAMP is able to find! There are two main parameters that influence this. The number of ambiguous bases allowed within a primer and the threshold for consensus nucleotides. Setting the threshold higher or the number of ambiguous bases lower will result in less conserved regions. If you have set the parameters below and get a decent output, increase the threshold until the output gets worse. This will increase the specificity of your primers. Likewise, if you do not have a good output, consider increasing the number of ambiguous bases before you lower the threshold. The console output varVAMP will also give you some suggestions.
3. **varVAMP reported primer dimers. What now?**
-In your case varVAMP could not find suitable replacement primers. You can either rerun varVAMP and try different settings or you can perform a third pool that contains a amplicon that has one of the conflicting dimers. Notably, varVAMP also reports the dimer melting temperature. If it is still reasonable low, using a hot start polymerase might still lead to successful PCR amplification.
+In your case varVAMP could not find suitable replacement primers in the TILED mode. You can either rerun varVAMP and try different settings or you can perform a third pool that contains a amplicon that has one of the conflicting dimers. Notably, varVAMP also reports the dimer melting temperature. If it is still reasonable low, using a hot start polymerase might still lead to successful PCR amplification.
+4. **I have multiple amplicons after SANGER mode. Which should I use?**
-4. **How fast is varVAMP?**
+varVAMP sorts all amplicons by score and always takes the best one of non-overlapping amplicons. If you are not interested in a specific gene region, amplicon_0 is your best candidate!
+
+5. **How fast is varVAMP?**
+
+varVAMP is pretty fast given the complexity of the problem. Running time is depended on the alignment length, number of sequences and the running mode. While the TILED is rather slow, qPCR and SANGER can be faster. An alignment with a few hundred sequences and with a genome size of 10 kb will likely run in under a minute for the TILED mode. For large e.g. DNA viruses (200 kb) it takes considerably longer, but should still finish in minutes. Running time optimizations are planned.
-varVAMP is pretty fast given the complexity of the problem. Running time is depended on the alignment length, number of sequences and the running mode. While the TILED is rather slow, qPCR and SANGER are faster. An alignment with a few hundred sequences and with a genome size of 10 kb will likely run in under a minute for the TILED mode. For large e.g. DNA viruses (200 kb) it takes considerably longer, but should still finish in minutes. Running time optimizations are planned.
diff --git a/docs/how_varvamp_works.md b/docs/how_varvamp_works.md
index 63a53dd..0493ae6 100644
--- a/docs/how_varvamp_works.md
+++ b/docs/how_varvamp_works.md
@@ -37,7 +37,7 @@ To search for the best scoring amplicon, varVAMP uses a graph based approach.
7. Lastly, the best scoring scheme is evaluated for primer dimers in their respective pools. If a primer dimer pair is found, varVAMP evaluates for each primer their overlapping previously not considered primers (primer search step 2) and again minimizes the score. The scheme and all primers are updated. If no alternative primers can be found, varVAMP issues a warning and reports the unsolvable primer dimers.
#### Sanger sequencing
-coming soon
+varVAMP sorts all amplicons by their score and takes the non-overlapping amplicon with the lowest score! As varVAMP gives a size penalty to amplicons, varVAMP automatically finds amplicons with low primer scores close to your optimal length (if possible).
#### qPCR
coming soon
diff --git a/docs/installation.md b/docs/installation.md
index b7ef2b6..b98a51d 100644
--- a/docs/installation.md
+++ b/docs/installation.md
@@ -3,10 +3,14 @@ varVAMP runs on UNIX Systems, MacOSX and Windows with Python3 >=3.9 installed.
## Installation
+From PyPI:
+
```shell
pip install varvamp
```
+
That was already it. To check if it worked:
+
```shell
varvamp -v
```
diff --git a/docs/preparing_the_data.md b/docs/preparing_the_data.md
index 9757aae..a850e3a 100644
--- a/docs/preparing_the_data.md
+++ b/docs/preparing_the_data.md
@@ -12,7 +12,7 @@ mafft my_sequences.fasta > my_alignment.fasta
### BUT...
-If your sequences are two diverse, also varVAMP will not perform well. Therefore, it is important that, phylogenically, the input alignment makes sense. To analyze this you can calculate a tree with tools like [iqtree](http://www.iqtree.org/) and then use [TreeCluster](https://github.com/niemasd/TreeCluster) to get phylogenically related sequence clusters. However, this can be also computationally intensive.
+If your sequences are too diverse, also varVAMP will not perform well. Therefore, it is important that, phylogenically, the input alignment makes sense. To analyze this you can calculate a tree with tools like [iqtree](http://www.iqtree.org/) and then use [TreeCluster](https://github.com/niemasd/TreeCluster) to get phylogenically related sequence clusters. However, this can be also computationally intensive.
We have had good experience in using varVAMP with the sequence identity based clustering algorithm [vsearch](https://github.com/torognes/vsearch). A good starting point is a sequence identity between 0.8 and 0.85. For such clusters varVAMP should perform reasonably well.
diff --git a/docs/usage.md b/docs/usage.md
index 723df01..e22cae2 100644
--- a/docs/usage.md
+++ b/docs/usage.md
@@ -15,34 +15,33 @@ In this case varVAMP uses as standard settings:
* ```MAX_LENGTH``` = 2000 (maximum amplicon length)
* ```OVERLAP``` = 100 (minimum overlap length)
* ```THRESHOLD``` = 0.9 (nucleotide consensus threshold)
-* ```ALLOWED_AMBIGUOUS``` = 4 (number of allowed ambiguous characters in a primer)
+* ```N_AMBIG``` = 4 (number of allowed ambiguous characters in a primer)
+* ```MODE``` = TILED
These settings are quite relaxed and can produce decent results for diverse viruses (80-90 % sequence identity). However, you can likely optimize the result.
**Full usage:**
```shell
-varvamp