|
1 | | -# Raptor [](https://github.com/seqan/raptor/actions) [](https://codecov.io/gh/seqan/raptor) |
| 1 | +# Raptor [![build status][1]][2] [![codecov][3]][4] [![install with bioconda][5]][6] |
| 2 | + |
| 3 | +[1]: https://github.com/seqan/raptor/workflows/Raptor%20CI/badge.svg?branch=master |
| 4 | +[2]: https://github.com/seqan/raptor/actions |
| 5 | +[3]: https://codecov.io/gh/seqan/raptor/branch/master/graph/badge.svg?token=SJVMYRUKW2 |
| 6 | +[4]: https://codecov.io/gh/seqan/raptor |
| 7 | +[5]: https://img.shields.io/badge/install%20with-bioconda-brightgreen.svg?style=flat |
| 8 | +[6]: https://bioconda.github.io/recipes/raptor/README.html |
| 9 | + |
2 | 10 | ### A fast and space-efficient pre-filter for querying very large collections of nucleotide sequences |
3 | 11 |
|
4 | 12 | ## Download and Installation |
5 | | -There may be performance benefits when compiling from source, especially when using `-march=native` as compiler |
6 | | -directive. |
| 13 | +There may be performance benefits when compiling from source as the build can be optimized for the host system. |
7 | 14 |
|
8 | 15 | ### Install with bioconda (Linux) |
9 | | -[](http://bioconda.github.io/recipes/raptor/README.html) |
10 | 16 |
|
11 | 17 | ```bash |
12 | 18 | conda install -c bioconda -c conda-forge raptor |
|
62 | 68 |
|
63 | 69 | The binary can be found in `bin`. |
64 | 70 |
|
65 | | -You may want to add the raptor executable yo your PATH: |
| 71 | +You may want to add the Raptor executable to your PATH: |
66 | 72 | ``` |
67 | 73 | export PATH=$(pwd)/bin:$PATH |
68 | 74 | raptor --version |
69 | 75 | ``` |
70 | 76 |
|
| 77 | +By default, Raptor will be built with host specific optimizations (`-march=native`). This behavior can be disabled by |
| 78 | +passing `-DRAPTOR_NATIVE_BUILD=OFF` to CMake. |
| 79 | + |
71 | 80 | </details> |
72 | 81 |
|
73 | 82 | ## Example Data and Usage |
74 | | -A toy data set can be found [here](https://ftp.imp.fu-berlin.de/pub/seiler/raptor/). |
| 83 | +A toy data set (124 MiB compressed, 983 MiB decompressed) can be found |
| 84 | +[here](https://ftp.imp.fu-berlin.de/pub/seiler/raptor/). |
75 | 85 |
|
76 | 86 | ```bash |
77 | 87 | wget https://ftp.imp.fu-berlin.de/pub/seiler/raptor/example_data.tar.gz |
@@ -113,7 +123,7 @@ Afterwards, we can search for some reads: |
113 | 123 | raptor search --error 2 --index raptor.index --query example_data/64/reads/mini.fastq --output search.output |
114 | 124 | ``` |
115 | 125 |
|
116 | | -The output starts with a header section (lines starting with `\#`). The header maps a number to each input file. |
| 126 | +The output starts with a header section (lines starting with `#`). The header maps a number to each input file. |
117 | 127 | After the header section, each line of the output consists of the read ID (in the toy example these are numbers) and |
118 | 128 | the corresponding bins in which they were found: |
119 | 129 | ```text |
@@ -169,6 +179,18 @@ The preprocessing applies the same cutoffs as used in Mantis |
169 | 179 | This means that only minimisers that occur more often than the cutoff specifies are included in the output. |
170 | 180 | If you wish to process all minimisers, you can use `--disable-cutoffs`. |
171 | 181 |
|
| 182 | +### Partitioned indices |
| 183 | +To reduce the overall memory consumption, the index can be divided into multiple (a power of two) parts. |
| 184 | +This can be done by passing `--parts n` to `raptor build`, where `n` is the number of parts you want to create. |
| 185 | +This will create `n` files, each representing one part of the index. The `--size` parameter describes the overall size |
| 186 | +of the index. For example, `--size 8g --parts 4` will create four 2 GiB indices. This will reduce the memory consumption |
| 187 | +of `raptor build` and `raptor search` by approximately 6 GiB, since there will only be one part in memory at any given |
| 188 | +time. `raptor search` will automatically detect the parts, and does not need any special parameters. |
| 189 | + |
| 190 | +### Upgrading the index (v1.1.0 to v2.0.0) |
| 191 | +An old index can be upgraded by running `raptor upgrade` and providing some information about how the index was |
| 192 | +constructed. |
| 193 | + |
172 | 194 | ### SOCKS interface |
173 | 195 | We implement the core interface of [SOCKS](https://gitlab.ub.uni-bielefeld.de/gi/socks). |
174 | 196 | For a list of options, see the help pages: |
|
0 commit comments