Merge commit 'a3686deab294cb1baa8c544024a5d65dcad90846'

soedinglab · May 30, 2024 · afa738b · afa738b
2 parents 528cddc + a3686de
commit afa738b
Show file tree

Hide file tree

Showing 79 changed files with 829 additions and 334 deletions.
diff --git a/lib/mmseqs/.cirrus.yml b/lib/mmseqs/.cirrus.yml
@@ -4,7 +4,7 @@ env:
 task:
   name: FreeBSD-13
   freebsd_instance:
-    image_family: freebsd-13-0
+    image_family: freebsd-13-2-snap
   install_script: pkg install -y cmake git samtools
   compile_script: |
     mkdir build && cd build

diff --git a/lib/mmseqs/cmake/MMseqsSetupDerivedTarget.cmake b/lib/mmseqs/cmake/MMseqsSetupDerivedTarget.cmake
@@ -1,12 +1,16 @@
 include(AppendTargetProperty)
 
 function (mmseqs_setup_derived_target TARGET)
-    get_target_property(COMPILE_TMP mmseqs-framework COMPILE_FLAGS)
-    get_target_property(LINK_TMP mmseqs-framework LINK_FLAGS)
-    get_target_property(DEF_TMP mmseqs-framework COMPILE_DEFINITIONS)
-    get_target_property(INCL_TMP mmseqs-framework INCLUDE_DIRECTORIES)
+    set(SOURCE "${ARGN}")
+    if(NOT SOURCE)
+        set(SOURCE "mmseqs-framework")
+    endif()
+    get_target_property(COMPILE_TMP ${SOURCE} COMPILE_FLAGS)
+    get_target_property(LINK_TMP ${SOURCE} LINK_FLAGS)
+    get_target_property(DEF_TMP ${SOURCE} COMPILE_DEFINITIONS)
+    get_target_property(INCL_TMP ${SOURCE} INCLUDE_DIRECTORIES)
 
-    target_link_libraries(${TARGET} mmseqs-framework)
+    target_link_libraries(${TARGET} ${SOURCE})
     append_target_property(${TARGET} COMPILE_FLAGS ${COMPILE_TMP})
     append_target_property(${TARGET} LINK_FLAGS ${LINK_TMP})
     set_property(TARGET ${TARGET} APPEND PROPERTY COMPILE_DEFINITIONS ${DEF_TMP})

diff --git a/lib/mmseqs/data/workflow/blastp.sh b/lib/mmseqs/data/workflow/blastp.sh
@@ -5,6 +5,33 @@ fail() {
     exit 1
 }
 
+abspath() {
+    if [ -d "$1" ]; then
+        (cd "$1"; pwd)
+    elif [ -f "$1" ]; then
+        if [ -z "${1##*/*}" ]; then
+            echo "$(cd "${1%/*}"; pwd)/${1##*/}"
+        else
+            echo "$(pwd)/$1"
+        fi
+    elif [ -d "$(dirname "$1")" ]; then
+        echo "$(cd "$(dirname "$1")"; pwd)/$(basename "$1")"
+    fi
+}
+
+fake_pref() {
+    QDB="$1"
+    TDB="$2"
+    RES="$3"
+    # create link to data file which contains a list of all targets that should be aligned
+    ln -s "$(abspath "${TDB}.index")" "${RES}"
+    # create new index repeatedly pointing to same entry
+    INDEX_SIZE="$(wc -c < "${TDB}.index")"
+    awk -v size="$INDEX_SIZE" '{ print $1"\t0\t"size; }' "${QDB}.index" > "${RES}.index"
+    # create dbtype (7)
+    awk 'BEGIN { printf("%c%c%c%c",7,0,0,0); exit; }' > "${RES}.dbtype"
+}
+
 notExists() {
 	[ ! -f "$1" ]
 }
@@ -27,14 +54,23 @@ ALN_RES_MERGE="$TMP_PATH/aln_0"
 while [ "$STEP" -lt "$STEPS" ]; do
     SENS_PARAM=SENSE_${STEP}
     eval SENS="\$$SENS_PARAM"
-    # call prefilter module
+
+    # 1. Prefilter hits
     if notExists "$TMP_PATH/pref_$STEP.dbtype"; then
-        # shellcheck disable=SC2086
-        $RUNNER "$MMSEQS" prefilter "$INPUT" "$TARGET" "$TMP_PATH/pref_$STEP" $PREFILTER_PAR -s "$SENS" \
-            || fail "Prefilter died"
+      if [ "$PREFMODE" = "EXHAUSTIVE" ]; then
+          fake_pref "${INPUT}" "${TARGET}" "$TMP_PATH/pref_$STEP"
+      elif [ "$PREFMODE" = "UNGAPPED" ]; then
+          # shellcheck disable=SC2086
+          $RUNNER "$MMSEQS" ungappedprefilter "$INPUT" "$TARGET" "$TMP_PATH/pref_$STEP" $UNGAPPEDPREFILTER_PAR \
+              || fail "Ungapped prefilter died"
+      else
+          # shellcheck disable=SC2086
+          $RUNNER "$MMSEQS" prefilter "$INPUT" "$TARGET" "$TMP_PATH/pref_$STEP" $PREFILTER_PAR -s "$SENS" \
+              || fail "Prefilter died"
+      fi
     fi
 
-    # call alignment module
+    # 2. alignment module
     if [ "$STEPS" -eq 1 ]; then
         if notExists "$3.dbtype"; then
             # shellcheck disable=SC2086

diff --git a/lib/mmseqs/data/workflow/blastpgp.sh b/lib/mmseqs/data/workflow/blastpgp.sh
@@ -5,6 +5,33 @@ fail() {
     exit 1
 }
 
+abspath() {
+    if [ -d "$1" ]; then
+        (cd "$1"; pwd)
+    elif [ -f "$1" ]; then
+        if [ -z "${1##*/*}" ]; then
+            echo "$(cd "${1%/*}"; pwd)/${1##*/}"
+        else
+            echo "$(pwd)/$1"
+        fi
+    elif [ -d "$(dirname "$1")" ]; then
+        echo "$(cd "$(dirname "$1")"; pwd)/$(basename "$1")"
+    fi
+}
+
+fake_pref() {
+    QDB="$1"
+    TDB="$2"
+    RES="$3"
+    # create link to data file which contains a list of all targets that should be aligned
+    ln -s "$(abspath "${TDB}.index")" "${RES}"
+    # create new index repeatedly pointing to same entry
+    INDEX_SIZE="$(wc -c < "${TDB}.index")"
+    awk -v size="$INDEX_SIZE" '{ print $1"\t0\t"size; }' "${QDB}.index" > "${RES}.index"
+    # create dbtype (7)
+    awk 'BEGIN { printf("%c%c%c%c",7,0,0,0); exit; }' > "${RES}.dbtype"
+}
+
 notExists() {
 	[ ! -f "$1" ]
 }
@@ -28,15 +55,26 @@ STEP=0
 while [ "$STEP" -lt "$NUM_IT" ]; do
     # call prefilter module
     if notExists "$TMP_PATH/pref_tmp_${STEP}.done"; then
-        PARAM="PREFILTER_PAR_$STEP"
-        eval TMP="\$$PARAM"
+        if [ "$PREFMODE" = "EXHAUSTIVE" ]; then
+            TMP=""
+            PREF="fake_pref"
+        elif [ "$PREFMODE" = "UNGAPPED" ]; then
+            PARAM="UNGAPPEDPREFILTER_PAR_$STEP"
+            eval TMP="\$$PARAM"
+            PREF="${MMSEQS} ungappedprefilter"
+        else
+            PARAM="PREFILTER_PAR_$STEP"
+            eval TMP="\$$PARAM"
+            PREF="${MMSEQS} prefilter"
+        fi
+
         if [ "$STEP" -eq 0 ]; then
             # shellcheck disable=SC2086
-            $RUNNER "$MMSEQS" prefilter "$QUERYDB" "$2" "$TMP_PATH/pref_$STEP" ${TMP} \
+            $RUNNER $PREF "$QUERYDB" "$2" "$TMP_PATH/pref_$STEP" ${TMP} \
                 || fail "Prefilter died"
         else
             # shellcheck disable=SC2086
-            $RUNNER "$MMSEQS" prefilter "$QUERYDB" "$2" "$TMP_PATH/pref_tmp_$STEP" ${TMP} \
+            $RUNNER $PREF "$QUERYDB" "$2" "$TMP_PATH/pref_tmp_$STEP" ${TMP} \
                 || fail "Prefilter died"
         fi
         touch "$TMP_PATH/pref_tmp_${STEP}.done"

diff --git a/lib/mmseqs/data/workflow/createtaxdb.sh b/lib/mmseqs/data/workflow/createtaxdb.sh
@@ -59,7 +59,7 @@ if { [ "${DBMODE}" = "1" ] && notExists "${TAXDBNAME}_taxonomy"; } || { [ "${DBM
         # Download NCBI taxon information
         if notExists "${TMP_PATH}/ncbi_download.complete"; then
             echo "Download taxdump.tar.gz"
-            downloadFile "https://ftp.ncbi.nih.gov/pub/taxonomy/taxdump.tar.gz" "${TMP_PATH}/taxdump.tar.gz"
+            downloadFile "https://ftp.ncbi.nlm.nih.gov/pub/taxonomy/taxdump.tar.gz" "${TMP_PATH}/taxdump.tar.gz"
             tar -C "${TMP_PATH}" -xzf "${TMP_PATH}/taxdump.tar.gz" names.dmp nodes.dmp merged.dmp delnodes.dmp
             touch "${TMP_PATH}/ncbi_download.complete"
             rm -f "${TMP_PATH}/taxdump.tar.gz"

diff --git a/lib/mmseqs/data/workflow/databases.sh b/lib/mmseqs/data/workflow/databases.sh
@@ -118,9 +118,9 @@ case "${SELECTION}" in
         if notExists "${TMP_PATH}/nr.gz"; then
             date "+%s" > "${TMP_PATH}/version"
             downloadFile "https://ftp.ncbi.nlm.nih.gov/blast/db/FASTA/nr.gz" "${TMP_PATH}/nr.gz"
-            downloadFile "https://ftp.ncbi.nih.gov/pub/taxonomy/accession2taxid/prot.accession2taxid.gz" "${TMP_PATH}/prot.accession2taxid.gz"
+            downloadFile "https://ftp.ncbi.nlm.nih.gov/pub/taxonomy/accession2taxid/prot.accession2taxid.gz" "${TMP_PATH}/prot.accession2taxid.gz"
             gunzip "${TMP_PATH}/prot.accession2taxid.gz"
-            downloadFile "https://ftp.ncbi.nih.gov/pub/taxonomy/accession2taxid/pdb.accession2taxid.gz" "${TMP_PATH}/pdb.accession2taxid.gz"
+            downloadFile "https://ftp.ncbi.nlm.nih.gov/pub/taxonomy/accession2taxid/pdb.accession2taxid.gz" "${TMP_PATH}/pdb.accession2taxid.gz"
             gunzip "${TMP_PATH}/pdb.accession2taxid.gz"
         fi
         push_back "${TMP_PATH}/nr.gz"
@@ -147,7 +147,7 @@ case "${SELECTION}" in
     "PDB")
         if notExists "${TMP_PATH}/pdb_seqres.txt.gz"; then
             date "+%s" > "${TMP_PATH}/version"
-            downloadFile "https://ftp.wwpdb.org/pub/pdb/derived_data/pdb_seqres.txt.gz" "${TMP_PATH}/pdb_seqres.txt.gz"
+            downloadFile "https://files.wwpdb.org/pub/pdb/derived_data/pdb_seqres.txt.gz" "${TMP_PATH}/pdb_seqres.txt.gz"
         fi
         push_back "${TMP_PATH}/pdb_seqres.txt.gz"
         INPUT_TYPE="FASTA_LIST"
@@ -212,8 +212,8 @@ case "${SELECTION}" in
     ;;
     "CDD")
         if notExists "${TMP_PATH}/msa.msa.gz"; then
-            downloadFile "https://ftp.ncbi.nih.gov/pub/mmdb/cdd/cdd.info" "${TMP_PATH}/version"
-            downloadFile "https://ftp.ncbi.nih.gov/pub/mmdb/cdd/fasta.tar.gz" "${TMP_PATH}/msa.tar.gz"
+            downloadFile "https://ftp.ncbi.nlm.nih.gov/pub/mmdb/cdd/cdd.info" "${TMP_PATH}/version"
+            downloadFile "https://ftp.ncbi.nlm.nih.gov/pub/mmdb/cdd/fasta.tar.gz" "${TMP_PATH}/msa.tar.gz"
         fi
         INPUT_TYPE="FASTA_MSA"
         SED_FIX_LOOKUP='s|\.FASTA||g'

diff --git a/lib/mmseqs/data/workflow/taxpercontig.sh b/lib/mmseqs/data/workflow/taxpercontig.sh
@@ -45,7 +45,11 @@ if [ -n "${ORF_FILTER}" ]; then
     fi
 
     if notExists "${TMP_PATH}/orfs_aln.list"; then
-        awk '$3 > 1 { print $1 }' "${TMP_PATH}/orfs_aln.index" > "${TMP_PATH}/orfs_aln.list"
+        # shellcheck disable=SC2086
+        "$MMSEQS" recoverlongestorf "${ORFS_DB}" "${TMP_PATH}/orfs_aln" "${TMP_PATH}/orfs_aln_recovered.list" ${THREADS_PAR} \
+            || fail "recoverlongestorf died"
+        awk '$3 > 1 { print $1 }' "${TMP_PATH}/orfs_aln.index" > "${TMP_PATH}/orfs_aln_remain.list"
+        cat  "${TMP_PATH}/orfs_aln_recovered.list" "${TMP_PATH}/orfs_aln_remain.list" > "${TMP_PATH}/orfs_aln.list"
     fi
 
     if notExists "${TMP_PATH}/orfs_filter.dbtype"; then

diff --git a/lib/mmseqs/data/workflow/tsv2exprofiledb.sh b/lib/mmseqs/data/workflow/tsv2exprofiledb.sh
@@ -22,19 +22,19 @@ fi
 
 if notExists "${OUT}.dbtype"; then
   "$MMSEQS" tsv2db "${IN}.tsv" "${OUT}_tmp" --output-dbtype 0 ${VERBOSITY}
-  MMSEQS_FOCE_MERGE=1 "$MMSEQS" compress "${OUT}_tmp" "${OUT}" ${VERBOSITY}
+  MMSEQS_FORCE_MERGE=1 "$MMSEQS" compress "${OUT}_tmp" "${OUT}" ${VERBOSITY}
   "$MMSEQS" rmdb "${OUT}_tmp" ${VERBOSITY}
 fi
 
 if notExists "${OUT}_seq.dbtype"; then
   "$MMSEQS" tsv2db "${IN}_seq.tsv" "${OUT}_seq_tmp" --output-dbtype 0 ${VERBOSITY}
-  MMSEQS_FOCE_MERGE=1 "$MMSEQS" compress "${OUT}_seq_tmp" "${OUT}_seq" ${VERBOSITY}
+  MMSEQS_FORCE_MERGE=1 "$MMSEQS" compress "${OUT}_seq_tmp" "${OUT}_seq" ${VERBOSITY}
   "$MMSEQS" rmdb "${OUT}_seq_tmp" ${VERBOSITY}
 fi
 
 if notExists "${OUT}_aln.dbtype"; then
   "$MMSEQS" tsv2db "${IN}_aln.tsv" "${OUT}_aln_tmp" --output-dbtype 5 ${VERBOSITY}
-  MMSEQS_FOCE_MERGE=1 "$MMSEQS" compress "${OUT}_aln_tmp" "${OUT}_aln" ${VERBOSITY}
+  MMSEQS_FORCE_MERGE=1 "$MMSEQS" compress "${OUT}_aln_tmp" "${OUT}_aln" ${VERBOSITY}
   "$MMSEQS" rmdb "${OUT}_aln_tmp" ${VERBOSITY}
 fi
 

diff --git a/lib/mmseqs/lib/alp/sls_pvalues.hpp b/lib/mmseqs/lib/alp/sls_pvalues.hpp
@@ -258,7 +258,7 @@ namespace Sls {
 			return rand_C;	
 		};
 
-		static inline double standard_normal()//generates standard normal random value using the Box–Muller transform
+		static inline double standard_normal()//generates standard normal random value using the Box-Muller transform
 		{
 			double r1=0;
 			while(r1==0)

diff --git a/lib/mmseqs/lib/ksw2/kseq.h b/lib/mmseqs/lib/ksw2/kseq.h
@@ -113,9 +113,10 @@ typedef struct __kstring_t {
 					if (ks->end == -1) { ks->is_eof = 1; return -3; }	\
 				} else break;											\
 			}															\
-			if (delimiter == KS_SEP_LINE) { \
-				for (i = ks->begin; i < ks->end; ++i) \
-					if (ks->buf[i] == '\n') { ks->newline+=(append == 1); break; } \
+			if (delimiter == KS_SEP_LINE) { 							\
+				unsigned char *sep = (unsigned char*)memchr(ks->buf + ks->begin, '\n', ks->end - ks->begin); \
+    			i = sep != NULL ? sep - (unsigned char*)ks->buf : ks->end; \
+    			ks->newline += (sep != NULL && append == 1); 				\
 			} else if (delimiter > KS_SEP_MAX) {						\
 				for (i = ks->begin; i < ks->end; ++i)					\
 					if (ks->buf[i] == delimiter) break;					\

diff --git a/lib/mmseqs/src/CMakeLists.txt b/lib/mmseqs/src/CMakeLists.txt
@@ -76,8 +76,8 @@ if (NOT EMSCRIPTEN)
 endif()
 
 if (ENABLE_WERROR)
-    append_target_property(mmseqs-framework COMPILE_FLAGS -Werror)
-    append_target_property(mmseqs-framework LINK_FLAGS -Werror)
+    append_target_property(mmseqs-framework COMPILE_FLAGS -Werror -Wno-unused-command-line-argument)
+    append_target_property(mmseqs-framework LINK_FLAGS -Werror -Wno-unused-command-line-argument)
 endif()
 
 # needed for concat.h
@@ -222,6 +222,7 @@ if (OPENMP_FOUND)
     if (NOT "${CMAKE_CXX_COMPILER_ID}" STREQUAL "GNU")
         target_link_libraries(mmseqs-framework ${OpenMP_CXX_LIBRARIES})
     endif()
+    target_include_directories(mmseqs-framework PUBLIC ${OpenMP_CXX_INCLUDE_DIRS})
     append_target_property(mmseqs-framework COMPILE_FLAGS ${OpenMP_CXX_FLAGS})
     append_target_property(mmseqs-framework LINK_FLAGS ${OpenMP_CXX_FLAGS})
 elseif (REQUIRE_OPENMP)

diff --git a/lib/mmseqs/src/CommandDeclarations.h b/lib/mmseqs/src/CommandDeclarations.h
@@ -23,6 +23,7 @@ extern int convertkb(int argc, const char **argv, const Command& command);
 extern int convertmsa(int argc, const char **argv, const Command& command);
 extern int convertprofiledb(int argc, const char **argv, const Command& command);
 extern int createdb(int argc, const char **argv, const Command& command);
+extern int makepaddedseqdb(int argc, const char **argv, const Command& command);
 extern int createindex(int argc, const char **argv, const Command& command);
 extern int createlinindex(int argc, const char **argv, const Command& command);
 extern int createseqfiledb(int argc, const char **argv, const Command& command);
@@ -97,6 +98,7 @@ extern int ungappedprefilter(int argc, const char **argv, const Command& command
 extern int gappedprefilter(int argc, const char **argv, const Command& command);
 extern int unpackdb(int argc, const char **argv, const Command& command);
 extern int rbh(int argc, const char **argv, const Command& command);
+extern int recoverlongestorf(int argc, const char **argv, const Command& command);
 extern int result2flat(int argc, const char **argv, const Command& command);
 extern int result2msa(int argc, const char **argv, const Command& command);
 extern int result2dnamsa(int argc, const char **argv, const Command& command);

diff --git a/lib/mmseqs/src/MMseqsBase.cpp b/lib/mmseqs/src/MMseqsBase.cpp
@@ -39,9 +39,9 @@ std::vector<Command> baseCommands = {
                 "Slower, sensitive clustering",
                 "mmseqs easy-cluster examples/DB.fasta result tmp\n"
                 "# Cluster output\n"
-                "#  - result_rep_seq.fasta: Representatives\n"
-                "#  - result_all_seq.fasta: FASTA-like per cluster\n"
-                "#  - result_cluster.tsv:   Adjacency list\n\n"
+                "#  - result_rep_seq.fasta:  Representatives\n"
+                "#  - result_all_seqs.fasta: FASTA-like per cluster\n"
+                "#  - result_cluster.tsv:    Adjacency list\n\n"
                 "# Important parameter: --min-seq-id, --cov-mode and -c \n"
                 "#                  --cov-mode \n"
                 "#                  0    1    2\n"
@@ -62,9 +62,9 @@ std::vector<Command> baseCommands = {
                 "Fast linear time cluster, less sensitive clustering",
                 "mmseqs easy-linclust examples/DB.fasta result tmp\n\n"
                 "# Linclust output\n"
-                "#  - result_rep_seq.fasta: Representatives\n"
-                "#  - result_all_seq.fasta: FASTA-like per cluster\n"
-                "#  - result_cluster.tsv:   Adjecency list\n\n"
+                "#  - result_rep_seq.fasta:  Representatives\n"
+                "#  - result_all_seqs.fasta: FASTA-like per cluster\n"
+                "#  - result_cluster.tsv:    Adjecency list\n\n"
                 "# Important parameter: --min-seq-id, --cov-mode and -c \n"
                 "#                  --cov-mode \n"
                 "#                  0    1    2\n"
@@ -130,14 +130,21 @@ std::vector<Command> baseCommands = {
                 "<i:fastaFile1[.gz|.bz2]> ... <i:fastaFileN[.gz|.bz2]>|<i:stdin> <o:sequenceDB>",
                 CITATION_MMSEQS2, {{"fast[a|q]File[.gz|bz2]|stdin", DbType::ACCESS_MODE_INPUT, DbType::NEED_DATA | DbType::VARIADIC, &DbValidator::flatfileStdinAndGeneric },
                                                            {"sequenceDB", DbType::ACCESS_MODE_OUTPUT, DbType::NEED_DATA, &DbValidator::flatfile }}},
+        {"makepaddedseqdb",               makepaddedseqdb,              &par.onlyverbosity,              COMMAND_HIDDEN,
+                "Generate a padded sequence DB",
+                "Generate a padded sequence DB",
+                "Martin Steinegger <[email protected]>",
+                "<i:sequenceDB> <o:sequenceDB>",
+                CITATION_MMSEQS2, {{"sequenceDB", DbType::ACCESS_MODE_INPUT, DbType::NEED_DATA|DbType::NEED_HEADER, &DbValidator::sequenceDb },
+                                          {"sequenceIndexDB", DbType::ACCESS_MODE_OUTPUT, DbType::NEED_DATA, &DbValidator::sequenceDb }}},
         {"appenddbtoindex",      appenddbtoindex,      &par.appenddbtoindex,      COMMAND_HIDDEN,
                 NULL,
                 NULL,
                 "Milot Mirdita <[email protected]>",
                 "<i:DB1> ... <i:DBN> <o:DB>",
                 CITATION_MMSEQS2, {{"DB", DbType::ACCESS_MODE_INPUT, DbType::NEED_DATA | DbType::VARIADIC, &DbValidator::allDb },
                                    {"DB", DbType::ACCESS_MODE_INPUT, DbType::NEED_DATA, &DbValidator::allDb }}},
-       {"indexdb",               indexdb,              &par.indexdb,              COMMAND_HIDDEN,
+        {"indexdb",               indexdb,              &par.indexdb,              COMMAND_HIDDEN,
                 NULL,
                 NULL,
                 "Martin Steinegger <[email protected]>",
@@ -555,7 +562,7 @@ std::vector<Command> baseCommands = {
                                                            {"resultDB", DbType::ACCESS_MODE_INPUT, DbType::NEED_DATA, &DbValidator::resultDb },
                                                            {"pvalDB", DbType::ACCESS_MODE_OUTPUT, DbType::NEED_DATA, &DbValidator::resultDb },
                                                            {"tmpDir", DbType::ACCESS_MODE_OUTPUT, DbType::NEED_DATA, &DbValidator::directory }}},
-        {"mergeresultsbyset",    mergeresultsbyset,    &par.threadsandcompression,COMMAND_MULTIHIT,
+        {"mergeresultsbyset",    mergeresultsbyset,    &par.mergeresultsbyset,    COMMAND_MULTIHIT,
                 "Merge results from multiple ORFs back to their respective contig",
                 NULL,
                 "Ruoshi Zhang, Clovis Norroy & Milot Mirdita <[email protected]>",
@@ -900,6 +907,14 @@ std::vector<Command> baseCommands = {
                 "Eli Levy Karin <[email protected]> ",
                 "<i:contigsSequenceDB> <i:extractedOrfsHeadersDB> <o:orfsAlignedToContigDB>",
                 CITATION_MMSEQS2, {{"",DbType::ACCESS_MODE_INPUT, DbType::NEED_DATA, NULL}}},
+        {"recoverlongestorf",    recoverlongestorf,    &par.onlythreads,          COMMAND_EXPERT,
+                "Recover longest ORF for taxonomy annotation after elimination",
+                NULL,
+                "Sung-eun Jang",
+                "<i:orfDB> <i:resultDB> <o:tsvFile>",
+                CITATION_MMSEQS2, {{"orfDB", DbType::ACCESS_MODE_INPUT, DbType::NEED_DATA, &DbValidator::sequenceDb},
+                                                           {"resultDB", DbType::ACCESS_MODE_INPUT, DbType::NEED_DATA, &DbValidator::resultDb},
+                                                           {"tsvFile", DbType::ACCESS_MODE_OUTPUT, DbType::NEED_DATA, &DbValidator::flatfile}}},
         {"reverseseq",          reverseseq,            &par.reverseseq,           COMMAND_SEQUENCE,
                 "Reverse (without complement) sequences",
                 NULL,

diff --git a/lib/mmseqs/src/alignment/Matcher.cpp b/lib/mmseqs/src/alignment/Matcher.cpp
@@ -282,7 +282,7 @@ size_t Matcher::resultToBuffer(char * buff1, const result_t &result, bool addBac
     *(tmpBuff-1) = '\t';
     tmpBuff = Util::fastSeqIdToBuffer(result.seqId, tmpBuff);
     *(tmpBuff-1) = '\t';
-    tmpBuff += sprintf(tmpBuff,"%.3E",result.eval);
+    tmpBuff += snprintf(tmpBuff, 32, "%.3E", result.eval);
     tmpBuff++;
     *(tmpBuff-1) = '\t';
     tmpBuff = Itoa::i32toa_sse2(result.qStartPos, tmpBuff);