Merge branch 'master' into v0.0

kdm9 · Aug 12, 2014 · 47d1c14 · 47d1c14
2 parents 2223a2a + 50311f9
commit 47d1c14
Show file tree

Hide file tree

Showing 3 changed files with 71 additions and 41 deletions.
diff --git a/README.md b/README.md
@@ -1,7 +1,7 @@
 AXE
 ===
 
-> Demultiplex NGS reads using fast data structures. It's fast, and made of tries!
+> Demultiplex NGS reads using trie data structures. It's fast, and made of tries!
 
 AXE very rapidly selects the optimal barcode present in a sequence read, even
 in the presence of sequencing errors. The algorithm is able to handle
@@ -17,17 +17,6 @@ existing demultiplexers. Unscientific trials show AXE processes more than
 **Warning**: Axe has not yet been comprehensively tested. However, in the
 spirit of "release early and often", here it is.
 
-Implementation Progress:
-------------------------
-
- - [x] Single ended read demultiplexing
- - [x] Interleaved/Paired input and output with single-ended demultiplexing
- - [x] Combinatorial demultiplexing
- - [x] CLI integration tests
- - [ ] Comprehensive `libaxe` tests
- - [ ] Comprehensive CLI tests
-
-See also TODO.md
 
 Installation:
 -------------
@@ -58,30 +47,57 @@ e.g.:
     cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=~/ ..
 
 For me, using `~/` as the prefix will install `axe` under `/home/kevin/bin` on
-GNU/Linux, and `/Users/kevin/bin` on Mac OSX.
-
-It's also wise to use `make install` not `sudo make install` when installing to
-a home directory.
+GNU/Linux, and (if I had one) `/Users/kevin/bin` on Mac OSX.It's also wise to
+use `make install` not `sudo make install` when installing to a home directory.
 
 ###Dependencies:
 
-- cmake. This is installable via `sudo apt-get install cmake` on debian based
+- cmake. This is installable via `sudo apt-get install cmake` on Debian based
   systems, or `brew install cmake` using homebrew on OS X.
-- zlib version >= 1.2.5. On debian, use the pacakge `zlib1g-dev`.
+- zlib version >= 1.2.5. On Debian, use the pacakge `zlib1g-dev`.
 - kmlib, tinytest and libdatrie (bundled in source, if you used
   `git clone --recursive` or an installation tarball)
 
+You'll possibly need to install zlib to your chosen prefix (e.g. `~/`) on
+supercomuters, which often have very old versions of zlib. To do so:
+
+    wget http://zlib.net/zlib-1.2.8.tar.gz
+    tar xvf zlib-1.2.8.tar.gz
+    cd zlib-1.2.8
+    ./configure --prefix=<your_prefix> # e.g. --prefix=~/
+    make && make install
+
+And then, use the following cmake command, assuming your prefix is `~/`:
+
+    cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=~/ -DZLIB_ROOT=~/ ..
+
+
 Usage:
 ------
 
 Full documentation, including a basic description of the algorithm, is hosted
 at https://axe-demultiplexer.readthedocs.org/en/latest/ .
 
+
+Implementation Progress:
+------------------------
+
+ - [x] Single ended read demultiplexing
+ - [x] Interleaved/Paired input and output with single-ended demultiplexing
+ - [x] Combinatorial demultiplexing
+ - [x] CLI integration tests
+ - [ ] Comprehensive `libaxe` tests
+ - [ ] Comprehensive CLI tests
+
+See also TODO.md
+
+
 Publication
 -----------
 
 A publication is coming soon, if the reviewer gods decide to smile upon us.
 
+
 LICENSE
 -------
 

diff --git a/src/axe.c b/src/axe.c
@@ -486,14 +486,16 @@ load_tries_combo(struct axe_config *config)
 
     if (!axe_config_ok(config)) {
         fprintf(stderr, "[load_tries] Bad config\n");
-        return -1;
+        ret = -1;
+        goto exit;
     }
     /* Make mutated barcodes and add to trie */
     for (iii = 0; iii < config->n_barcode_pairs; iii++) {
         this_bcd = config->barcodes[iii];
         if (!axe_barcode_ok(this_bcd)) {
             fprintf(stderr, "[load_tries] Bad R1 barcode at %zu\n", iii);
-            return -1;
+            ret = -1;
+            goto exit;
         }
         /* Either lookup the index of the first read in the barcode table, or
          * insert this barcode into the table, storing its index.
@@ -503,7 +505,8 @@ load_tries_combo(struct axe_config *config)
             if (ret != 0) {
                 fprintf(stderr, "ERROR: Could not load barcode %s into trie %zu\n",
                         this_bcd->seq1, iii);
-                return 1;
+                ret = 1;
+                goto exit;
             }
         } else {
             continue;
@@ -512,7 +515,10 @@ load_tries_combo(struct axe_config *config)
             /* Do the forwards read barcode */
             mutated = hamming_mutate_dna(&num_mutated, this_bcd->seq1,
                                          this_bcd->len1, jjj, 0);
-            assert(mutated != NULL);
+            if (mutated == NULL) {
+                ret = 1;
+                goto exit;
+            }
             for (mmm = 0; mmm < num_mutated; mmm++) {
                 ret = axe_trie_add(config->fwd_tries[jjj], mutated[mmm], bcd1);
                 if (ret != 0) {
@@ -521,7 +527,6 @@ load_tries_combo(struct axe_config *config)
                             fprintf(stderr,
                                     "[%s] warning: Will only match %s to %dmm\n",
                                     __func__, this_bcd->id, (int)jjj - 1);
-
                         }
                         trie_delete(config->fwd_tries[jjj]->trie,
                                     mutated[mmm]);
@@ -548,7 +553,8 @@ load_tries_combo(struct axe_config *config)
             if (ret != 0) {
                 fprintf(stderr, "ERROR: Could not load barcode %s into trie %zu\n",
                         this_bcd->seq2, iii);
-                return 1;
+                retval = 1;
+                goto exit;
             }
         } else {
             continue;
@@ -557,7 +563,10 @@ load_tries_combo(struct axe_config *config)
             num_mutated = 0;
             mutated = hamming_mutate_dna(&num_mutated, this_bcd->seq2,
                                          this_bcd->len2, jjj, 0);
-            assert(mutated != NULL);
+            if (mutated == NULL) {
+                ret = 1;
+                goto exit;
+            }
             for (mmm = 0; mmm < num_mutated; mmm++) {
                 ret = axe_trie_add(config->rev_tries[jjj], mutated[mmm], bcd2);
                 if (ret != 0) {
@@ -566,7 +575,6 @@ load_tries_combo(struct axe_config *config)
                             fprintf(stderr,
                                     "[%s] warning: Will only match %s to %dmm\n",
                                     __func__, this_bcd->id, (int)jjj - 1);
-
                         }
                         trie_delete(config->rev_tries[jjj]->trie,
                                     mutated[mmm]);
@@ -586,6 +594,7 @@ load_tries_combo(struct axe_config *config)
     }
     /* we got here, so we succeeded. set retval accordingly */
     retval = 0;
+
 exit:
     if (mutated != NULL) {
         for (mmm = 0; mmm < num_mutated; mmm++) {
@@ -638,7 +647,10 @@ load_tries_single(struct axe_config *config)
         for (jjj = 1; jjj <= config->mismatches; jjj++) {
             mutated = hamming_mutate_dna(&num_mutated, this_bcd->seq1,
                                          this_bcd->len1, jjj, 0);
-            assert(mutated != NULL);
+            if (mutated == NULL) {
+                ret = 1;
+                goto exit;
+            }
             for (mmm = 0; mmm < num_mutated; mmm++) {
                 ret = axe_trie_add(config->fwd_tries[jjj], mutated[mmm], iii);
                 if (ret != 0) {
@@ -647,7 +659,6 @@ load_tries_single(struct axe_config *config)
                             fprintf(stderr,
                                     "[%s] warning: Will only match %s to %dmm\n",
                                     __func__, this_bcd->id, (int)jjj - 1);
-
                         }
                         trie_delete(config->fwd_tries[jjj]->trie,
                                     mutated[mmm]);
@@ -668,11 +679,14 @@ load_tries_single(struct axe_config *config)
     }
     /* we got here, so we succeeded */
     retval = 0;
+
 exit:
-    for (mmm = 0; mmm < num_mutated; mmm++) {
-        km_free(mutated[mmm]);
+    if (mutated != NULL) {
+        for (mmm = 0; mmm < num_mutated; mmm++) {
+            km_free(mutated[mmm]);
+        }
+        km_free(mutated);
     }
-    km_free(mutated);
     return retval;
 }
 

diff --git a/src/main.c b/src/main.c
@@ -245,9 +245,6 @@ parse_args(struct axe_config *config, int argc, char * const *argv)
                 fprintf(stderr, "ERROR: Setting revese read input file failed.\n");
                 goto error;
                 break;
-                fprintf(stderr, "ERROR: Setting interleaved input file failed.\n");
-                goto error;
-                break;
             case READS_UNKNOWN:
             default:
                 goto error;
@@ -264,6 +261,9 @@ parse_args(struct axe_config *config, int argc, char * const *argv)
                 goto error;
                 break;
             case READS_SINGLE:
+                fprintf(stderr, "ERROR: Revese read input file set in single-end mode.\n");
+                goto error;
+                break;
             case READS_UNKNOWN:
             default:
                 /* Misc weirdness */
@@ -369,16 +369,16 @@ parse_args(struct axe_config *config, int argc, char * const *argv)
     }
     config->have_cli_opts = 1;
     return 0;
-version:
-    axe_config_destroy(config);
-    print_version();
-    return 0;
-help:
-    config->have_cli_opts = 0;
-    return 2;
 error:
     config->have_cli_opts = 0;
     return 1;
+help:
+    config->have_cli_opts = 0;
+    return 2;
+version:
+    print_version();
+    axe_config_destroy(config);
+    exit(0);
 }
 
 int